Search

The Multi-Objective Discrete Network Design Problem (MO-DNDP) describes the optimisation of a transport network across multiple metrics. Due to the impractical processing times of exact solutions, heuristics have been harnessed to solve this problem. However, the scale of the networks, the set of candidate links and the metrics assessed have been limited. This research proposes a MO-DNDP that encapsulates 8 metrics across three real networks used as benchmarks scaling from 17 km2 to 41,850 km2 to encapsulate the complexity of optimising transport networks at different scales, traffic demands and constraint densities. To solve this more comprehensive MO-DNDP, this paper proposes the Physarum-inspired algorithm PANDA that improves upon the classical Physarum Solver algorithm by harnessing the attractant sensing and merging capabilities of the biological Physarum Polycephalum slime mould. PANDA is a first attempt at solving the complex MO-DNDP using the Physarum foraging model. PANDA consistently scored better than the classical Solver in the land use rating by circumventing landforms, built-up areas, and environmentally sensitive regions. Additionally, PANDA constructed networks that were on average 30 % better than the existing benchmark networks and 32 % better than the Physarum Solver networks. PANDA improved upon the benchmarks by constructing lower capacity roads that were optimised for the actual traffic demand between each origin–destination pair instead of shared arterials. [ABSTRACT FROM AUTHOR]

High dimensional and small samples characterize gene expression data and contain a large number of genes unrelated to disease. Feature selection improves the efficiency of disease diagnosis by selecting a small number of important genes. Unfortunately, existing algorithms do not consider the correlation between features, and search algorithms tend to fall into the local optimal solution in the feature search process. To this end, this paper proposes a feature filter and group evolution hybrid feature selection algorithm (FG-HFS) for high-dimensional gene expression data. Unlike existing algorithms, we propose using spectral clustering to group redundant features into a group. Then, we propose a redundant feature filter algorithm. According to the principle of approximate Markov blanket, grouped feature groups are filtered to delete these redundant features. Among them, filtered features are evenly divided by density according to the feature exponential strategy. Most importantly, we propose using the group evolution multi-objective genetic algorithm to search the filtered feature subsets and evaluate the candidate feature subsets according to the in-group and out-group so as to select the feature subsets with the highest accuracy and the least number. Experimental results show that the average accuracy (ACC) and Matthews correlation coefficient (MCC) indexes of the selected feature subsets (FSs) by the FG-HFS algorithm on 5 gene expression datasets are 92.76% and 88.76%, respectively, which are significantly better than the existing algorithms. In addition, the FSs and ACC/FSs indexes of the FG-HFS algorithm are also better than the existing algorithms, which fully proves the superiority of the FG-HFS algorithm. More importantly, the Wilcoxon and Friedman statistical experiments results show that the feature selection effect of FG-HFS algorithm is significantly better than that of existing algorithms, no matter in pairwise comparison or multiple comparison. • We propose using spectral clustering to group the features so that the in-group feature similarity is extremely high. • We propose a redundant feature filter algorithm since existing algorithms cannot filter redundant features. • we propose using the group evolution multi-objective genetic algorithm to search the filtered feature subsets. [ABSTRACT FROM AUTHOR]

This paper introduces a novel surrogate-assisted multi-objective nutcracker optimization algorithm. This algorithm is built upon the recently proposed nutcracker optimization algorithm, drawing inspiration from the behaviours observed in Clark's nutcrackers. The algorithm is developed based on two distinct behaviours exhibited by these birds. To comprehensively evaluate the performance of the proposed algorithm, a dual-pronged approach is adopted. On the one hand, a set of artificial test problems is employed to scrutinize the algorithm's capabilities, while on the other hand, a set of real-world problems is considered to assess its practical efficacy. The results of the proposed algorithm are evaluated in comparison to existing baseline algorithms and state-of-the-art algorithms, using well-recognized performance metrics, both qualitatively and quantitatively. The obtained results provide convincing evidence of the performance of the proposed algorithm. [ABSTRACT FROM AUTHOR]

This paper focuses on the deadlock control and scheduling problem of two-stage assembly permutation flowshop with limited buffers. In such a system, all parts are first processed in a permutation flowshop, then assembled into final products. The objective is to find a feasible schedule, including a manufacturing subschedule and an assembly subschedule, to minimize makespan. Owing to limited buffers, not only may the blocking phenomenon appear in the system operation, but also some manufacturing subschedules may lead to deadlock states, making them unavailable. To obtain a feasible schedule, this work first proposes a deadlock control or an availability detection and amendment (ADA) algorithm to determine whether a manufacturing subschedule is available, and convert unavailable ones to available ones. Then by embedding ADA into a social spider algorithm (SSA), a hybrid social spider algorithm (HSSA) is developed for computing feasible schedules. In HSSA, an improved heuristic search algorithm is proposed to generate a better initial population, and a simplified variable neighborhood search algorithm is designed and incorporated into HSSA to enhance its searching ability. Finally, HSSA is tested on instances, showing its superiority over SSA in performance. The effect of buffer size on the scheduling objective is discussed. • The availability of manufacturing subschedules is determined. • The unavailable subschedules can be transformed into available ones. • A hybrid social spider algorithm is developed for computing feasible schedules. • Using an improved heuristic search algorithm to obtain a better initial population. • The local search ability is enhanced by a simplified variable neighborhood search. [ABSTRACT FROM AUTHOR]

• This paper proposes a new DOSbFC algorithm to remove baseline drift of EOG Signals. • This paper presents a new electrode positioning scheme based on eyeglasses. • This paper provides a long-term eye movement detection function with high accuracy. This paper presents a new method to remove baseline drift and noise by using a differential electrooculography (EOG) signal based on a fixation curve (DOSbFC) and a new electrode positioning scheme based on eyeglasses for user convenience. In addition, a desktop application and mobile applications to control the human–computer interface were implemented. Finally, we created experimental EOG eyeglasses and a new detection protocol using the proposed method for long-term step-by-step detection of eye movements and user comfort. The proposed DOSbFC calculates the difference values of accumulated EOG signals between the initial eye movement and fixation time. It allows long-term detection of eye movements with high accuracy and only requires a single calibration. The vertical and ground electrodes of the standard electrode positioning scheme caused discomfort of subjects; the proposed electrode positioning scheme solves these problems and enables the use of existing eyeglasses without design modification. The experimental results demonstrated that the average accuracy of the long-term eye movement detection was 94%, whereas those of the band pass filter and wavelet transform were 61% and 64%, respectively. This was because baseline drift and noise were removed by averaging the signal variations. Further experimental results demonstrated that the average information transfer rate of the proposed method was 6.0, whereas those of the band pass filter and wavelet transform were 1.1 and 0.9, respectively. [ABSTRACT FROM AUTHOR]

The fuzzy c-means (FCM) clustering algorithm requires the pre-definition of the number of clusters. When used for imbalanced datasets, FCM tends to equalize the sizes of clusters and thus produces bad clustering results. Although clustering validity index (CVI) is commonly employed to determine the number of clusters, it is often ineffective for bad clustering results of FCM. To address this issue, we have previously proposed a fuzzy CVI, namely IMI, for imbalanced datasets. However, IMI encounters issues when the number of clusters exceeds three, as it tends to divide the majority clusters into multiple subclusters. In this paper, a novel CVI, IMI2, is proposed to enhance the performance of IMI. IMI2 adopts a two-step clustering algorithm that merges some clusters from the IMI result based on their separation values. This enhancement overcomes the limitations of IMI, rendering IMI2 an excellent CVI capable of accurately determining the number of clusters for datasets with multiple imbalanced clusters. The performance of IMI2 is evaluated using several synthetic and UCI datasets. The results demonstrate that IMI2 outperforms other methods in terms of its superior performance for multiple imbalanced clusters. IMI2 retains the good performance of IMI for datasets with two or three clusters, while significantly improving it for multiple imbalanced clusters. • We propose a new clustering validity index that is effective for unbalanced data sets with multiple clusters. • We design a merging algorithm that is able to produce majority clusters of high quality. • This new CVI performs better than existing CVIs and some clustering algorithms. [ABSTRACT FROM AUTHOR]

Resistor network research is of great importance, yet many resistor networks and their large-scale fast computations have not received sufficient attention. This paper proposes a new resistor network with idiosyncratic shape, i.e., a α × β conic surface (CS) resistor network that resembles the upper part of a three-dimensional Dirac function. Utilizing the Recursion Transform (RT-V) method of Tan, a recursive matrix equation model is constructed based on Kirchhoff's law and nodal voltages, which contains the modified tridiagonal Toeplitz matrix. By using the orthogonal matrix transformation, the eigenvalues and eigenvectors of the modified tridiagonal Toeplitz are obtained. The discrete sine transform of the fourth type (DST − IV) is utilized to solve node voltages, while the explicit potential function is represented by the Chebyshev polynomials of the second kind. In addition, explicit potential functions for some special cases are provided, and the potential distribution is illustrated using dynamic three-dimensional graph. To achieve a rapid calculation of the potential, a fast algorithm based on the multiplication of DST-IV with a vector is proposed. In the end, analysis of computational efficiency for the explicit potential function and the fast algorithm are shown. • A new conic surface resistor network is established. • The explicit potential function of represented by Chebyshev polynomials. • The potential in special cases are displayed by a three-dimensional dynamic view. • A fast algorithm of computing potential is proposed by using famous DST-IV. • Experimental analysis of computational efficiency and computational scale. [ABSTRACT FROM AUTHOR]

Addressing the challenges of low convergence accuracy and stagnation at local optima in the application of the golden jackal optimizer (GJO) to mobile robot path planning, this paper proposes a hybrid strategy-based golden jackal optimizer (HGJO) algorithm. The improved algorithm employs a pre-decreasing slow nonlinear energy decay strategy to balance the global and local search capabilities. The roulette wheel selection algorithm and Lévy flight strategy are introduced into the position update of the GJO algorithm, so the proposed algorithm avoids stagnation at the local optimum. The HGJO algorithm is evaluated against some state-of-the-art optimizers on 23 benchmark functions and the CEC2021 benchmark function. It is also applied to ablation experiments for mobile robot path planning. The experimental results show that the HGJO algorithm improves the average path length in path planning by 0.21%, 82.4%, and 7.9% over the original algorithm in three different environments under 30 independent experiments. • A new energy decreasing method is proposed. • A roulette wheel selection strategy is introduced. • A new hybrid position updating strategy is proposed. [ABSTRACT FROM AUTHOR]

In texture classification, local binary pattern (LBP) is currently one of the most widely-concerned feature encoding models. Most existing LBP-based texture classification methods are usually limited to single-kind texture features. In fact, an across-domain fusion of LBP features with other features, such as image contours, could be another potential path to promote texture classification. To enhance the feature modeling ability of LBP-based methods, this paper firstly designs a Cellular Neural Network (CellNN) with recurrent convolutions, initially trained by a simplified simulated-annealing algorithm, to extract informative image contours. For better reliability, a new three-channel contour extractor of deep CellNNs (i. e. , dCellNNs) is proposed. This extractor contains the initially-trained CellNNs of more than three layers, and it is further optimized by fine-tuning parameters. Moreover, a new weighted-base algorithm is designed to fulfill the fusion of the multi-scale texture features by LBPs and the contour features by dCellNNs to enhance feature representation. Finally, these enhanced features are concatenated together to generate the final multi-scale features of given texture image. On texture datasets KTH, Brodatz, OTC12 and UIUC, experiment results verify that the across-domain fusion of multi-scale LBPs and dCellNNs is efficient in capturing & enhancing texture features. With moderate feature dimensionality and computational costs, it could improve texture classification, acquiring an obvious accuracy increase on previous state-of-the-art ones, e.g. , a rise of 2.58% on KTH-TIPS2b, a rise of 3.11% on Brodatz, a rise of 0.71% on OTC12 and a rise of 0.42% on UIUC. • A multi-scale across-domain feature fusion of textures and contours is proposed. • The first deep model of CellNN (dCellNNs) with two-stage training is proposed. • The three-channel texture contour extractor of dCellNNs is designed. • A weighted-base fusing is designed for the across-domain fusion of features. [ABSTRACT FROM AUTHOR]

We investigate a multi-stage newsvendor problem with advance purchase discount (APD) in this paper. At the beginning of a stage, the decision maker (DM) makes the advance ordering decision for all the periods in this stage; at the start of every period within the stage, the DM makes the regular ordering decision. In this problem, the only available information we can observe is the past demands. To solve this problem, we extend the weak aggregating algorithm (WAA) with one decision variable, an online learning approach based on the theory of prediction and learning with expert advice, to a two-dimensional problem that involves advance ordering decisions in stages and regular ordering decisions nested in each stage. The difficulty of the problem lies in transferring learned knowledge of demand information from stage to stage. We design a cross-stage knowledge transfer scheme and obtain online ordering solutions for both advance-order and regular-order. We show that our solutions converge to the optimal solutions asymptotically. In addition, we derive theoretical guarantees for total gains in one stage and cumulative gains for all stages in the planning horizon. Through numerical studies, we find that our solutions are competitive to those offered by the best experts in hindsight. Finally, we do the sensitivity analysis to illustrate the effectiveness of our algorithm under different parameter values. • Consider advance purchase discounts in a multi-stage newsvendor problem. • Extend weak aggregating algorithm from one- to a two-dimensional case. • Obtain analytical solutions for both advance order and regular order. • Derive theoretical guarantees for one stage and all stages' profits. [ABSTRACT FROM AUTHOR]

Optimal path planning is a fundamental problem in artificial intelligence (AI) and has wide applications in areas such as robotics, transportation, and logistics. In this paper, we propose a novel approach for ordering algorithms in pathfinding problems using the concept of shell layers. Our approach aims to improve the computational efficiency and distance cost of ordering algorithms. To evaluate the effectiveness of our approach, we compare it with five state-of-the-art techniques commonly used in the field of ordering algorithms on eight different scenarios with varying configurations. Our results show that our proposed approach outperformed the state-of-the-art techniques in terms of time-computational complexity and distance cost in most of the scenarios, demonstrating its potential as a new state-of-the-art technique for ordering algorithms. Specifically, our approach outperformed the nearest neighbor, A*, branch and bound, Christofides, and genetics ordering algorithms. However, we also identified two specific situations where our approach was outperformed by two state-of-the-art algorithms due to the specific distribution of the goal nodes. These findings highlight the importance of evaluating and comparing ordering algorithms in various scenarios. Our approach has significant implications for AI research and development, as it has the potential to improve the performance of pathfinding algorithms in various applications. We also discuss the limitations of our approach and potential areas for further research, such as investigating the effectiveness of our approach in different types of graphs and exploring the potential use of machine learning techniques to optimize the shell layer construction. The code for this study is available at https://github.com/abdullah1aloush1/MuGONA. • Equally-spaced nodes formed over workspace with obstacles, optimizing search process. • Efficient neighboring graph built for equally-spaced nodes, reducing complexity. • Path optimization for navigation by ordering goal nodes, reducing time and distance. • Path smoothing for navigation via interpolation to yield a trajectory on path nodes. [ABSTRACT FROM AUTHOR]

A vision system comprises several steps, with each step exerting a significant impact on the final outcome. One of these crucial steps is segmentation, which isolates the region of interest within an object and removes the background. Segmentation is vital because it enhances the quality of the isolated region, improves the accuracy of extracted features, and reduces the noise introduced by poor-quality features. Several segmentation techniques are available in the literature, each requiring one or more adjustable parameters. Selecting the most appropriate technique for segmenting a particular image dataset can be challenging. Several factors can affect the segmentation quality, including preprocessing, the chosen segmentation method, parameter fine-tuning approaches, and the characteristics of the dataset. Moreover, variations in lighting and intensity can further influence segmentation quality. At times, an expert may need to manually choose the segmentation technique and fine-tune its associated parameters. This paper presents the development of an automated algorithm for the selection of segmentation techniques and their associated parameters. The developed techniques are implemented and compared using diverse datasets, and the resulting experimental outcomes are thoroughly discussed and analyzed. The algorithm aims to streamline and simplify the process of selecting appropriate segmentation techniques, determining the required parameters, and selecting suitable pre-processing techniques. • New method to segmentation of image data-sets. • The proposed algorithm prevents poor segmentation due to bright changes. • The proposed algorithm improves the quality of the data-set segmentation. • Automatically found best parameters and space color to segmentation. • We compare the results using PRI VOI and GCE metrics. [ABSTRACT FROM AUTHOR]

The influence maximization problem asks to find a small number of early adopters of a product in a social network, such that the expected number of total adoptions is maximized over the network. The problem has been well-studied, but most of the studies focus on the case of a single product or purely competitive products. This paper proposes a new influence diffusion model for multiple complementary products, namely, the comparative linear threshold (Com-LT) model. Under the Com-LT model, we model the complementary relation by reducing the thresholds of nodes. With this model, we study two problems: SelfInfMax and CompInfMax. We prove that these two problems are both NP-Hard under the Com-LT model. For both the SelfInfMax and the CompInfMax problem, we theoretically analyze the monotonicity and submodularity, and accordingly leverage lower bound optimization to devise non-trivial effective approximation algorithms. We conduct experiments over 4 real-world datasets. The experimental results demonstrate the correctness and efficiency of the proposed algorithms. • We study the complementary version of the influence maximization problem. • We propose Com-LT model to describe the influence spread of non- competitive products. • We studied the SelfInfMax and the CompInfMax problem and show their hardness. • We propose approximation algorithms with high quality feasible solutions. • We conduct experiments on real networks and show the effectiveness of our algorithms. [ABSTRACT FROM AUTHOR]

Stochastic optimization (SO) algorithms based on the Powerball function, namely powered stochastic optimization (PoweredSO) algorithms, have been confirmed, effectively, and demonstrated great potential in the context of large-scale optimization and machine learning tasks. Nevertheless, the issue of how to determine the learning rate for PoweredSO is a challenge and still unsolved problem. In this paper, we propose a class of adaptive PoweredSO approaches that are efficient, scalable and robust. It takes advantage of the hypergradient descent (HD) technique to automatically acquire an online learning rate for PoweredSO-like methods. In the first part, we study the behavior of the canonical PoweredSO algorithm, the Powerball stochastic gradient descent (pbSGD) method, with HD. The existing PoweredSO algorithms also suffer from the high variance because they take the similar algorithmic framework to SO algorithms, arising from sampling tactics. Therefore, the second portion develops an adaptive powered variance-reduced optimization method via utilizing both variance-reduced technique and HD. Moreover, we present the convergence analysis of the proposed algorithms and explore their iteration complexity on non-convex cases. Numerical experiments are conducted on machine learning tasks, verifying the superior performance over modern SO algorithms. • We propose a type of adaptive powered stochastic optimization algorithms. • The theoretical analysis of the resulting algorithms is established. • Experiments on machine learning tasks verify the efficacy of the algorithms. [ABSTRACT FROM AUTHOR]

Predictor-based Neural Architecture Search is an important topic since it can efficiently reduce the computational cost of evaluating candidate architectures. Most existing predictor-based NAS algorithms aim to design different predictors to improve prediction performance. Unfortunately, even a promising performance predictor may suffer from the accuracy decline due to long-term and continuous usage, thus leading to the degraded performance of the search strategy. That naturally gives rise to the following problems: how do predictors affect search strategies and how to efficiently use the predictor? In this paper, we take Reinforcement Learning (RL) based search strategy to study theoretically and empirically the impact of predictors on search strategies. We first formulate an RL-Predictor-based NAS algorithm as model-based RL and analyze it with a guarantee of monotonic improvement. Then, based on this analysis, we propose a simple procedure of predictor usage, named m i x e d b a t c h , which contains ground-truth data and prediction data in a batch. The proposed procedure can efficiently reduce the impact of predictor errors on the RL-based search strategy with maintaining performance growth. Our algorithm, RL-Predictor-based Neural Architecture Search with Mixed batch (RPNASM), outperforms traditional NAS algorithms and prior state-of-the-art predictor-based NAS algorithms on three NAS-Bench-201 tasks and one NAS-Bench-ASR task. Our code is available at https://github.com/tjdeng/RPNASM. • Theoretically analyze predictor's impact on RL search strategy for the first time. • Perform comprehensive experiments to investigate RL-Predictor based NAS algorithms. • Propose RL-Predictor-based NAS framework to enhance search performance. [ABSTRACT FROM AUTHOR]

In recent years, the flexible job shop dynamic scheduling problem (FJDSP) has received considerable attention; however, FJDSP with transportation resource constraint is seldom investigated. In this study, FJDSP with transportation resource constraint is considered and an improved non-dominated sorting genetic algorithm-III (NSGA-III) algorithm (DQNSGA) integrated with reinforcement learning (RL) is proposed. In DQNSGA, an initialization method based on heuristic rules and an insertional greedy decoding approach are designed, and a double-Q Learning with an improved ε -greedy strategy is used to adaptively adjust the key parameters of NSGA-III. An improved elite selection strategy is also applied. Through extensive experiments and practical case studies, this algorithm has been compared with three other well-known algorithms. The results demonstrate that DQNSGA exhibits significant effectiveness and superiority in all tests. The research presented in this paper enables effective adjustments of production plans in response to dynamic events, which is of critical importance for production management in the manufacturing industry. [ABSTRACT FROM AUTHOR]

Prediction-based dynamic multiobjective optimization evolutionary algorithms have become one of the mainstream methods for solving dynamic multiobjective optimization problems. However, the unknown nonlinear relationships in sequential environments bring great challenge to the construction of prediction model and the inevitable error between the historical obtained solutions and the real pareto-optimal solutions troubles the prediction accuracy. In this paper, a gradient boosting regression tree based dynamic multiobjective optimization evolutionary algorithm is proposed, called MOEA/D-XGB. The time series solutions formed by subspace decomposition are trained in the XGBoost-based predictor to produce a high-quality initial population when environmental change occurs. And a novel population promotion strategy based on generalized additive model is proposed to improve the quality of historical obtained solutions. The performance comparisons with five state-of-the-art algorithms have shown that MOEA/D-XGB achieves best performance in 59 out of 70 experiments on MIGD and 60 out of 70 experiments on MHV, respectively, which demonstrates that the proposed design is capable of significantly improving the performance of dynamic multiobjective optimization. [ABSTRACT FROM AUTHOR]

Recently, an image encryption algorithm using Quantum Chaotic Map and DNA Coding (QCMDC-IEA) has been reported. It consists of two main components: pixel-level permutation and DNA domain substitution. To support its ability to withstand various attacks, several security analyses and experimental simulations were presented. However, after careful cryptanalysis, we found that QCMDC-IEA has inherent fatal security problems. Although formally using complex chaos and DNA encoding, the chaos-based sequences used for encryption in QCMDC-IEA are independent of a plain image such that it suffers from the defect of the existence of an equivalent key. Moreover, the lack of confusion and diffusion in DNA domain encryption makes it vulnerable to cryptographic attacks. DNA domain encryption is then essentially a 2-bit data substitution process, so it can be equivalently simplified. On this basis, we propose an attack method to crack QCMDC-IEA, which first obtains an equivalent permutation key by differential cryptanalysis, and then eliminates the DNA domain substitution based on a chosen-plaintext attack using only four special plain images and their corresponding cipher images, and finally recovers the original plain images. Our attack method takes full advantage of the security defects in QCMDC-IEA and achieves complete decipherment with low complexity, thus better revealing its intrinsic security mechanism. To improve the security performance, some security enhancement suggestions are recommended for similar cryptosystems. Both theoretical analysis and experimental simulation results show that the proposed cryptographic attack method is effective and feasible for QCMDC-IEA with low attack complexity. Therefore, the cryptanalysis work in this paper can provide some theoretical hints for improving the security of a class of image encryption algorithms based on DNA coding and chaos. [ABSTRACT FROM AUTHOR]

The mining of frequent itemsets and association rules is a core problem in data mining and an essential task in data analysis. In this paper, we present SufRec, a new algorithm for finding frequent itemsets and association rules. We give two versions. In both versions, the mining of the frequent itemsets is decomposed as a sequence of tasks. The first version (the SufRecDep algorithm) proceeds the tasks successively, each one using the results of the previous ones. The second version (the SufRecInd algorithm) performs the tasks independently of each other. Both versions are recursive with respect to the items and have thus the advantage to be particularly efficient for updating the mining process when new items are added to the database or when others are excluded. Moreover, the task-independent processing of SufRecInd makes it very easy to build parallel versions of SufRec. We present two parallel SufRec algorithms. With the first one, P-FIFO-SufRec, the processors perform the tasks according to the order in which the items appear in the database, preserving the recursive nature of SufRec. For second one, P-PAST-SufRec, the set of items is fixed and the tasks are pre-assigned to each processor. In order to evaluate the performance of SufRec, we carry out a consequent experimental study. In particular, we compare its running times with those of well-known parallel and non-parallel frequent itemsets mining algorithms. • Construction of SufRec, a new algorithm for frequent itemsets mining (FIM). • Construction of parallel versions of SufRec algorithm. • Experimental study comparing SufRec with well-known FIM algorithms. • Applications to the case where the list of items changes over time. [ABSTRACT FROM AUTHOR]

Artificial bee colony (ABC) is a prominent algorithm that offers great exploration capabilities among various meta-heuristic algorithms. However, its monotonous and one-dimensional search strategy limits its searching performance in the solving process. Thus, to address this issue, a Q-learning based multi-strategy integrated ABC algorithm (QMABC) is proposed. In the QMABC, multiple search strategies are proposed to utilize different individual experiences and search approaches for solution updates. Then, Q-learning is employed for strategy selection. In comparison to previous studies, this paper introduces more effective state and action configurations within the framework of Q-learning. To evaluate the performance of the QMABC, CEC 2017 benchmark functions are adopted to compare it to different meta-heuristic algorithms including ABC based and non-ABC based algorithms. Moreover, applications in path planning are implemented to further verify the effectiveness of the QMABC. Overall, it should be highlighted that the proposed QMABC demonstrates superiority in both numerical and practical experiments. • A Q-learning based multi-strategy integrated ABC algorithm is proposed. • Multiple strategies are introduced to the onlooker bee phase. • Q-learning is utilized to adaptively determine the most suitable strategy. • Novel state and action settings are designed in Q-learning framework. • Proposed algorithm shows superiority in numerical and practical experiments. [ABSTRACT FROM AUTHOR]

To alleviate the challenges posed by privacy protection, sparse data, cold-start problem and weak interpretation, in this paper, we investigate how to uncover the implicit "social relationships" for user/item pairs and create recommendation relationships between users and items in the recommendation systems with no explicit trust relationships. Firstly, according to complex network theory, a recommendation system is presented by a heterogeneous network with five various types of subnets. Secondly, a novel algorithm ISimRank-ISR is designed based on the improved SimRank similarity index, which mainly integrates common and non-common behavior information, as well as attribute information for exploring the implicit social relations, and generates two types of social subnets. Further, introducing these implicit social relations into the nonnegative matrix factorization (NMF) model, another novel algorithm INMF-RRM is proposed to reconstruct the user-item rating subnet. Then, the integration of the above two algorithms derives an algorithm ISR-RRM to create recommendation relationships for the recommendation task. Finally, numerical experiments on two Movielens datasets are conducted to verify the feasibility and effectiveness of the three new algorithms mentioned above, and we come to the following conclusions. (1) The ISimRank-ISR algorithm provides an in-depth exploration of the implicit social relationships for user/item pairs. (2) The INMF-RRM algorithm has convergence properties. (3) To perform better and obtain more accurate recommendations, with the ISR-RRM algorithm, there are optimal parameters achieved. (4) Compared with the other seven benchmark algorithms, the ISR-RRM algorithm shows excellent performance, and its RMSE and MAE are reduced by a maximum of 14.75% and 8.05%, respectively. Additionally, it is more effective in alleviating the cold-start problem. [ABSTRACT FROM AUTHOR]

Feature extraction networks play a crucial role in classification algorithms. However, most convolutional neural networks use the form of residual block stacking to extract downstream features of images, and this simple form is not sufficient to obtain better classification performance. Therefore, this paper proposes a YOLOX detection framework based on an improved ConvNeXt which incorporates many advantages of Swin-Transformer with efficient classification performance. To address the problems of unstable IOU convergence and inaccurate regression coordinates of the algorithm, this paper adopts Manhattan distance as the new penalty term, and the new loss function combines the advantages of Euclidean distance and Manhattan distance, which has no gradient explosion in the early training period and can drop smoothly to a stable point in the late training period. The proposed algorithm is robust to negative effects such as dust, light, shadows, and occlusions, and is well-suited for masked-face detection in industrial and public scenarios. The proposed model is tested on the VOC dataset and the AIZOO open-source masked-face dataset, obtaining 5.58% AP improvement on the VOC dataset and reaching 96.665% AP on the masked-face dataset, which is better than other algorithms. • Designing a deep neural network to use for masked-face detection. • Anchor-free model is used to avoid time-consuming post-processing. • The improved ConvNeXt network improves classification performance. • Adopting Manhattan distance-based localization loss makes training stable. • The Mosaic data augmentation strategy alleviates the overfitting problem. [ABSTRACT FROM AUTHOR]

In this paper, we proposed a novel fitness-value-independent Differential Evolution to tackle complex real-parameter single-objective optimization. There were three innovations in the algorithm: First, dimension improvements based adaptation schemes for control parameters F and C R were advanced in this paper. Different from those fitness value improvements based parameter control in the recent winner DE variants, our algorithm has fitness-value-independent characteristic, therefore, it can be applied in much wider optimization scenarios especially for those that the exact fitness values of the objectives are unavailable. Second, a combined parabolic–linear reduction scheme for population size is employed in the algorithm based on the observation that the slower reduction of population size at the earlier stage of the evolution usually helps to get better perception of the objectives while a linear reduction of population size in the later stage helps to get better exploitation. Third, an indicator is proposed to monitor the diversity of the population and a corresponding population enhancement technique is launched when the diversity is detected bad. The first two innovations have the P arameter a daptive characteristic of the DE algorithm while the third innovation refers to p opulation e nhancement t echnique, therefore, we name our algorithm "the PaDE-pet algorithm". Then, the algorithm is validated under a larger test suite containing 88 benchmarks from CEC2013, CEC2014 and CEC2017 test suites for real-parameter single objective optimization, which may avoid over-fitting problem in comparison with employing a test suite containing a small number of benchmarks. The experiments support the superiority of the novel PaDE-pet algorithm in comparison with several recently proposed powerful DE variants. • A black-box model reflecting complex optimization without available fitness values. • The model reveals the fitness-value-dependent weakness of recent winner DE variants. • A novel fitness-value-independent DE algorithm was proposed to tackle the model. • New adaptation schemes for control parameters and population enhancement technique. • To avoid over-fitting, a larger test suite is employed in algorithm validation. [ABSTRACT FROM AUTHOR]

Highlights • Fault prediction improves the effectiveness of software quality assurance activities. • This paper focuses on building an effective fault prediction classifier. • Fault prediction model using Iterated feature selection algorithms with L-RNN. • We perform experiments on 19 open source projects. • Fault prediction model is best suitable for projects with faulty classes less than the threshold value. Abstract Software fault prediction (SFP) is typically used to predict faults in software components. Machine learning techniques (e.g., classification) are widely used to tackle this problem. With the availability of the huge amount of data that can be obtained from mining software historical repositories, it becomes possible to have some features (metrics) that are not correlated with the faults, which consequently mislead the learning algorithm and thus decrease its performance. One possible solution to eliminate those metrics is Feature Selection (FS). In this paper, a novel FS approach is proposed to enhance the performance of a layered recurrent neural network (L-RNN), which is used as a classification technique for the SFP problem. Three different wrapper FS algorithms (i.e, Binary Genetic Algorithm (BGA), Binary Particle Swarm Optimization (BPSO), and Binary Ant Colony Optimization (BACO)) were employed iteratively. To assess the performance of the proposed approach, 19 real-world software projects from PROMISE repository are investigated and the experimental results are discussed. Receiver operating characteristic - area under the curve (ROC-AUC) is used as a performance measure. The results are compared with other state-of-art approaches including Naïve Bayes (NB), Artificial Neural Network (ANN), logistic regression (LR), the k-nearest neighbors (k-NN) and C4.5 decision trees, in terms of area under the curve (AUC). Our results have demonstrated that the proposed approach can outperform other existing methods. [ABSTRACT FROM AUTHOR]

Highlights • A novel modeling strategy for Decision Trees is proposed. • Interactive basis functions induce oblique and non-linear decision boundaries. • Benchmarking shows potential gains in accuracy and/or parsimony of models. • Empirical examples illustrate the application of the modeling strategy. • Decision charts are introduced to improve the interpretability of models. Abstract Decision Trees (DTs) are a machine learning technique widely used for regression and classification purposes. Conventionally, the decision boundaries of Decision Trees are orthogonal to the features under consideration. A well-known limitation of this is that the algorithm may fail to find optimal partitions, or in some cases any partitions at all, depending on the underlying distribution of the data. To remedy this limitation, several modifications have been proposed that allow for oblique decision boundaries. The objective of this paper is to propose a new strategy for generating flexible decision boundaries by means of interactive basis functions (IBFs). We show how oblique decision boundaries can be obtained as a particular case of IBFs, and in addition how non-linear decision boundaries can be induced. One attractive aspect of the strategy proposed in this paper is that training Decision Trees with IBFs does not require custom software, since the functions can be precalculated for use in any existing implementation of the algorithm. Since the underlying mechanisms remain unchanged there is no substantial computational overhead compared to conventional trees. Furthermore, this also means that IBFs can be used in any extensions of the Decision Tree algorithm, such as evolutionary trees, boosting, and bagging. We conduct a benchmarking exercise to understand under which conditions the use of IBFs can improve model the performance. In addition, we present three empirical applications that illustrate the approach in classification and regression. As part of discussing the empirical applications, we introduce a device called decision charts to facilitate the interpretation of DTs with IBFs. Finally, we conclude the paper by outlining some directions for future research. [ABSTRACT FROM AUTHOR]

Highlights • Process model for transformation of formalized sentences into relational database. • Algorithms of functions and procedures for each step of transformation. • Meta model for the structure and content of formalized sentences in the relational database. Abstract Verbalized text contains knowledge necessary and sufficient for transfer of numerous human cognitions. The question is how this knowledge was saved into text. Authors believe that they found an important idea how to assemble knowledge into nodes of knowledge. Similar methods in the field of knowledge networks exist, but none of them does it in the same way as the Node of Knowledge (NOK) method. Basic terms and concepts in a language are represented in words whose meaning is final and cannot be divided in subterms. More complex meaning can be achieved by combining words in sentences. According to authors' opinion, each sentence includes connective medium for words in the sentence (related to semantic reasons), which is inwrought in the Node of Knowledge method. Using the Node of Knowledge method each sentence can be presented as a network of connected words. This network is enriched with links between words so a computer can interpret meaning and knowledge of the sentence in the same way an intelligent person does. A formalized and semantically enriched record of sentences (called Formalized Node of Knowledge – FNOK) is developed. Authors find that in this way, even without statistical text analysis, an algorithm can give correct answer to a question set based on written text. This paper presents the system for transformation of textual knowledge expressed in natural language sentences into a relational database. The system is a part of a larger knowledge-based system based on the Node of Knowledge (NOK) conceptual framework for knowledge-based system development. This paper starts with sentences written in the formalized and enriched form for which a logical transformation into the structure of a relational database is proposed. The system and algorithms for the transformation of formalized sentences into n-tuples of the relational database are distributed and represented in several steps. The research has shown that it is possible to store semantically enriched sentences in relational databases. The solution presented in this paper is important for further development of the system for receiving questions from users and providing answers (i.e. question-answering system), with the ability to use well-developed relational SQL languages. Relational database of texts enables numerous applications in the field of expert and intelligent systems. [ABSTRACT FROM AUTHOR]

This paper characterizes a class of repeated moral hazard problem where the agent's replacement action is necessary for assuring the performance of important materials and equipment. However, due to the principal's costly monitoring, the continuously decreasing performance caused by the agent's successive shirking behaviors will significantly harm the principal's profits or welfare. For sake of performance assurance and monitoring efficiency, this paper adopts two reinforcement learning algorithms to find and test the strategies on the final path satisfying the conditions of (Restricted) Experience Based Equilibrium. Experimental results in a case of quality assurance of grain demonstrate the effectiveness of the model and the solution methods. Practically, this framework especially fits a kind of government-enterprise relation, such as in the field of emergency materials management and safety production management. Given the increasing need of developing systems for the government's management, this paper aims to provide applicable strategies and guidelines from a theoretical perspective. [ABSTRACT FROM AUTHOR]

In complex networks, there are a huge number of overlapping community structures, and those structures have been an increasing concern in recent years. The label propagation algorithm (LPA) family is a kind of mainstream method that discovers different community structures in complex networks. But these methods often struggle with stability and feasibility, as well as issues sensitive to network size. In this paper, to solve the above problems and improve the detection accuracy, we present a fast label propagation algorithm (FLPA) based on node influence and label weight. In FLPA, we first use the graph compression technique to reduce the network size. Then, we propose a new node influence calculation method and fuse it with the α -path similarity to precisely control the label weight in the label propagation stage. Through the above process, FLPA makes the labels that are propagated and received by nodes more reasonable and further improves the detection accuracy. Finally, we restore the network and make the compressed node belong to the same community as its corresponding super node. Experimental results on 10 real and 58 synthetic networks indicate that FLPA is suitable for detecting overlapping community structures regardless of network scale and accomplishes better than state-of-the-art methods on stability and feasibility. [ABSTRACT FROM AUTHOR]

The continuous malicious software (malware) attacks on smartphones pose a serious threat to the security of users, especially the dominant platform Android. Data-driven methods based on machine learning algorithms are a promising way to defend against that. In this paper, we explore the limitations of this kind of methods in improving the performance of malware detection, e.g., considering each feature in isolation and relying on balanced class data, and propose a multi-model ensemble framework MEFDroid by combining individual predictors, where hybrid deep learning based feature extraction methods are adopted to learn meaningful features from raw data. Besides, a novel fusion scheme is exploited to fuse multi-level representations and mine their correlations to maximize the utilization of original information. To evaluate the effectiveness of the MEFDroid framework, we conduct a step-by-step model justification experiment to investigate how the proposed algorithms (i.e., ESAES, EDAES and EDAFS) enhance the malware detection performance gradually. We also compare the proposed algorithms with classical machine learning methods, conventional sampling solutions for the imbalance problem. Besides, we investigate the reliability of our proposed methods using another public dataset. Our extensive experimental results demonstrate that the target model EDAFS in MEFDroid outperforms others in terms of most metrics, which means that it is an effective alternative solution to detect Android malware. [ABSTRACT FROM AUTHOR]

One driving application for multi-robots is source seeking, especially in the hazardous environment. It consists of two essential subtasks: source location and path search. Nature inspired meta-heuristic is preferable in addressing the source location subtask which is an inversion problem, while the Astar algorithm and its variants are widely used for the path search subtask. In this paper, we present a multi-robot team simulator as a container which contains both algorithms as components. The simulator takes the constraints into consideration, including the size and the speed bound of each robot, the obstacle and collision avoidance. We provide a python implementation and example problems for research and test purposes. The well-structured code with object-oriented design can be conveniently upgraded by adding new excellent nature inspired metaheuristics, or extended to other source seeking problems in various field applications. The python code can be downloaded from the website: https://github.com/buctlab/source-seeking-multi-robot-team-simulator. • We developed a source-seeking multi-robot simulator containing many metaheuristics. • We implemented the multimodal version of metaheuristics in the simulator. • The simulator closely combines the metaheuristics with the Astar algorithm. [ABSTRACT FROM AUTHOR]

In this paper, a metaheuristic algorithm called PID-based search algorithm (PSA) is proposed for global optimization. The algorithm is based on an incremental PID algorithm that converges the entire population to an optimal state by continuously adjusting the system deviations. PSA is mathematically modeled and implemented to achieve optimization in a wide range of search spaces. PSA is used to solve CEC2017 benchmark test functions and six constrained problems. The optimization performance of PSA is verified by comparing it with seven metaheuristics proposed in recent years. The Kruskal-Wallis, Holm and Friedman tests verified the superiority of PSA in terms of statistical significance. The results show that PSA can be better balanced exploration and exploitation with strong optimization capability. Source codes of PSA are publicly available at https://ww2.mathworks.cn/matlabcentral/fileexchange/131534-pid-based-search-algorithm. [ABSTRACT FROM AUTHOR]

Most image encryption techniques are proposed to be applied to the whole image, which lacks the focus of information value. In this paper, we propose a target-based image encryption scheme. For remotely sensed airport images, we first take the aircraft considered to be sensitive information as the target and detect its positions. And then perform the proposed three-dimensional bit disorder method (3DBDM) and the arbitrary point dynamic diffusion on these regions. Finally, encrypt the entire image through high-speed index-dynamic diffusion to improve efficiency. To obtain a more stable key stream, a new ill-conditioned parameter infinite-interval chaotic system (IPIS) is introduced. The evaluation of IPIS shows that the introduction of ill-conditioned matrix brings higher sensitivity, randomness, and wider chaotic interval, which potentially improves the security of chaotic image encryption. In particular, the encryption method for aircraft regions can be generalized to ordinary images. Various encryption effect tests verify the security and practicability of the proposed algorithm and target-based encryption scheme. [ABSTRACT FROM AUTHOR]

The last few years have seen an increasing demand for high-capacity Internet services, and this need has intensified in the years 2020 and 2021. In 2020 and 2021, Internet usage grew by 50% in several European countries, mainly due to home office, video streaming, hybrid teaching, and others. High-capacity optical networks usually meet this growing demand for Internet services. Thus, investigations that can improve the quality of optical networks are highly relevant in the current context. One of the research problems in this area is related to the physical topology design (PTD) of optical networks, which is classified as NP-hard. Several studies on PTD consider the application of meta-heuristics that obtain suboptimal solutions in a time compatible with engineering applications. However, meta-heuristics and local search techniques have been combined in several other optimization problems, which is not typical for the PTD problem. This paper proposes a solution to the PTD problem that combines a known multipurpose optimization algorithm, the NSGA-III, with operators considering the domain-specific knowledge of the problem to provide superior-quality networks. According to our results, the new proposal presents quality up to 8% higher than previous proposals concerning the hypervolume metrics (HV), maintaining a similar computational cost. • Proposal of pivot rules for local search in many objective algorithms for PTD. • Provide a method for efficient exploration of neighborhood for network optimization. • Proposal of a new local search that considers both convergence and diversity. • Hybridization of NSGA-III with the proposed LS, providing superior quality solutions. [ABSTRACT FROM AUTHOR]

This paper introduces a novel hybrid binary variant of the Grey Wolf Optimizer (GWO) and the Naked Mole-Rat Algorithm (NMRA), called HbGNMR. The primary objective of this hybridization is to overcome the poor exploration and local optima stagnation problems of the basic NMRA as well as the poor exploitation of GWO. HbGNMR incorporates the global search strategy of GWO into the worker phase to enhance exploration, and a mating factor based on simulated annealing-inertia weight is added to the breeder phase to improve exploitation. Moreover, fourteen new time-varying binary transfer functions of two families: seven S-shaped and seven V-shaped functions have been analysed. Among these transfer functions, T V V 5 was identified as the most effective in converting solutions from continuous search space to binary search space and maintaining a balance between exploration and exploitation phases. The hybrid performance of HbGNMR is evaluated on twenty-three CEC'05, thirty CEC'17, and eighty CEC'21 benchmark functions. The statistical results demonstrate that HbGNMR outperforms nine well-known state-of-the-art algorithms according to Friedman-test and Wilcoxon rank-sum test. To test the effectiveness of the proposed algorithm for real-world applications, HbGNMR is employed to optimize a large dimensional planar-monopole ultrawideband (UWB) fragment-type antenna structure of an 18 × 18 binary matrix cells. Experimental testing of the prototype is conducted on a vector-network-analyser-based experimental-testbed, which shows good agreement with the simulated results. Furthermore, the optimized structure achieves large bandwidth (3.1-12.6 GHz), excellent gain, and high directivity for the 10 × 15 mm 2 compact size UWB antenna geometry. Therefore, the optimized antenna can be utilized for modern wireless applications. • Developed a novel binary optimization algorithm named HbGNMR. • Demonstrate impact of new time-varying S- and V-shaped transfer functions on HbGNMR. • The algorithm is successfully tested on CEC'05, CEC'17 and CEC'21 benchmark functions. • HbGNMR performs excellently in solving fragment-type UWB real-world antenna problem. [ABSTRACT FROM AUTHOR]

This research paper focuses on quaternion neural networks (QNNs) - a type of neural network wherein the weights, biases, and input values are all represented as quaternion numbers. Previous studies have shown that QNNs outperform real-valued neural networks in basic tasks and have potential in high-dimensional problem spaces. However, research on QNNs has been fragmented, with contributions from different mathematical and engineering domains leading to unintentional overlap in QNN literature. This work aims to unify existing research by evaluating four distinct QNN backpropagation algorithms, including the novel GHR-calculus backpropagation algorithm, and providing concise, scalable implementations of each algorithm using a modern compiled programming language. Additionally, the authors apply a robust Design of Experiments (DoE) methodology to compare the accuracy and runtime of each algorithm. The experiments demonstrate that the Clifford Multilayer Perceptron (CMLP) learning algorithm results in statistically significant improvements in network test set accuracy while maintaining comparable runtime performance to the other three algorithms in four distinct regression tasks. By unifying existing research and comparing different QNN training algorithms, this work develops a state-of-the-art baseline and provides important insights into the potential of QNNs for solving high-dimensional problems. • Clifford Multilayer Perceptrons outperform other quaternion neural network methods. • Multilayer GHR Calc. networks approximate real nonlinear functions reasonably well. • Classic QNN backpropagation techniques are outdated and should be avoided. [ABSTRACT FROM AUTHOR]

Outlier detection is one of the hot research in data mining, and has been applied to various fields such as network anomaly detection, image abnormal analysis, etc. In recent years, many outlier detecting algorithms have been proposed. However, these outlier detecting algorithms are hard to effectively detect global outliers, local outliers and outlier clusters at the same time. In this paper, we propose a novel outlier detecting algorithm based on the following ideas: (1) the density distribution should not be changed dramatically on local area; (2) the ratio of the number of k nearest neighbors and the number of reverse k nearest neighbors should not be very big. Based on above ideas, the proposed algorithm aims to find outlier turning points, then regards all outlier turning points and its sparse neighbors as outliers. Furthermore, the proposed algorithm use natural neighbors to obtain the neighborhood parameter k adaptively. The formal analysis and extensive experiments demonstrate that this technique can detect global outliers, local outliers and outlier clusters without neighborhood parameter k. • OTF Detect local outliers and outlier clusters simultaneously. • OTF does not need to set neighborhood parameter k artificially. • The distribution of density will not change dramatically. • OTF aim to find outlier turning points, then detect outliers and outlier clusters. [ABSTRACT FROM AUTHOR]

Palmprints have gained enormous popularity in the biometric industry recently. Due to the large amount of information provided by the palmprints and their forensic value, many large-scale identification systems are employing high resolution palmprint-based biometric systems with minutiae as the primary identification feature. Minutiae are the palm ridge endings or bifurcations that are represented by their location and orientation, i.e., (x , y , θ). Due to the large size of palmprints, approximately 1000 minutiae are extracted per palmprint on average. Besides the large number, another problem with the extracted minutiae is that a considerable amount is false. Besides increasing the number of minutiae matches between palmprints, these false minutiae reduce matching accuracy. In order to reduce minutiae matches and eliminate false minutia, previous studies have adopted multifaceted approaches such as associating a quality or confidence factor with each minutia, image registration, or inventing computationally efficient minutia descriptors. However, all these approaches require associating each minutia with additional parameters. In this paper, we propose a simple and intuitive histogram-based minutiae selection algorithm (MSA) using only the basic properties, i.e., (x , y , θ) to shortlist a subset of best minutiae candidates for matching. This provides the dual benefit of: (1) a reduced number of minutiae matches between palmprints, and (2) improved matching accuracy through the elimination of false minutiae. The proposed method does not require estimating any additional parameters for minutiae and is independent of minutia encoding and matching methods used. Our results are acquired on a popular and challenging high resolution palmprint dataset and show the efficacy of the proposed algorithm through a clear improvement in matching accuracy and a significant reduction in the number of required minutiae matches between palmprints. • Minutiae selection algorithm to pick best candidate minutiae for palmprint matching. • MSA reduces minutiae matches required and improves matching accuracy. • MSA is independent of minutiae encoding and matching methods. • MSA does not require prior image alignment or image registration between palmprints. • MSA is suitable for application at both offline and online stages. [ABSTRACT FROM AUTHOR]

Portfolio optimization is an essential issue in quantitative investing, which aims to find the best set of portfolios by allocating the proportion of assets. One of the most widely studied portfolio optimization models is the cardinality constrained mean–variance model, which incorporates real-world constraints on the number of selected assets and lower and upper bounds on the proportion of each asset. This paper presents a novel metaheuristic algorithm based on the Mayfly algorithm to solve the cardinality constrained mean–variance portfolio optimization problem. To better adapt to this problem, we design and introduce some new features to the proposed algorithm, including (1) a new cardinality constraint handling strategy; (2) a new local search strategy; and (3) changes to the crossover operator. We have designed comparison experiments for the proposed metaheuristic and evaluated its performance using five commonly used performance metrics. The experimental results show that the proposed approach achieves competitive performance on datasets of different sizes. The results also demonstrate the feasibility of this approach in solving the cardinality constrained mean–variance portfolio optimization problem. • The proposed algorithm is used for Cardinality Constrained Portfolio Optimization. • A fast cardinality constraint handling strategy. • A nuptial dance and bi-direction-based local search strategy. • Excellent comparison experimental results on benchmark from OR-Lib. [ABSTRACT FROM AUTHOR]

When planning maintenance tasks involving the switching off of parts of the network, electricity transmission operators must evaluate and take into account the associated risks. The French operator, RTE, has developed a three-step approach in which they first calculate the risks in every possible scenario, then plan the maintenance tasks by optimizing some risk measures, and finally validate the plan. In this paper, we have addressed the second step, developing mathematical models that approximate the original problem as well as a matheuristic algorithm. It is composed of three phases, combining integer linear models in the first phase with metaheuristics such as Variable Neighbourhood Descent and Path Relinking in the second and third phases. The computational results obtained on the benchmarks provided by the company show that our algorithm improves on the best reported solutions for most of the test instances. • We have developed a matheuristic algorithm for a real maintenance planning problem. • The algorithm combines integer linear models with GRASP, VND, and Path Relinking. • The solutions are refined in a final intensification phase. • Our algorithm improves on the best reported solutions for most of the instances. [ABSTRACT FROM AUTHOR]

In this paper we solve a practical variant of an arc routing problem. We target the close enough model in which clients can be served from relatively close arcs. This variant, known as the profitable close-enough arc routing problem, models real situations, such as inventory management or automated meter reading. We propose a heuristic based on the variable neighborhood search methodology to maximize the sum of profits of the clients served (penalized with the distance traveled). Our method incorporates efficient search strategies to speed up the optimization process, as required in practical applications. We present extensive experimentation over a benchmark of previously reported instances. Specifically, we first set the key search parameters of our method, and then compare it with the state-of-the-art heuristics for this problem. Our heuristic outperforms the previous algorithms published for this problem, as confirmed by the statistical analysis, which permits to draw significant conclusions. • A literature overview on close-enough is carried out. • A MultiStart-VNS algorithm is proposed and adapted to solve the PCEARP. • The experimental results proved the efficiency of the proposed algorithm. • The metaheuristic algorithm outperforms other existing methods for the problem. • The method has a great potential to be applied in emerging real-life problems. [ABSTRACT FROM AUTHOR]

In this study, an energy efficiency study has been carried out through the moment values of a quadrotor applied in the position control with the differential evolution algorithm (DE), which is under the evolutionary algorithms among meta-heuristic algorithms particle swarm optimization (PSO) which is among swarm-based algorithms and gravity search algorithm (GSA) among physics-based algorithms and charged system search (CSS) algorithm, which are all well-known in the literature. The efficient use of energy in electrically powered flying mechanisms is important in terms of completing its task. The aim of this study is to obtain proportional–integral–derivative (PID) coefficients that minimize the energy consumption caused by quadrotor moments. First of all, 12 PID coefficients calculated with traditional Ziegler Nicholes, and trying-error methods for the control of the quadrotor have been trained with DE, PSO, GSA and CSS algorithms and their cost values have been found for this problem. In the study, optimum PID coefficients have been found separately with each algorithm and torque data giving energy efficiency have been obtained by applying them to the state equations. Although the cost minimization convergence value obtained from DE does not look good compared to other algorithms, it has been observed that the moment values of the obtained PID coefficients in the control program are lower. In the study, it has been shown that the PID coefficients obtained from DE reach the target with approximately less energy than the maximum-moment value of the PSO, which provides the closest moment value. It can be thought that this value contributes to energy consumption in each orbital movement of the quadrotor. The simulation results show that the coefficients obtained from DE provide better energy efficiency and are superior to other algorithms compared in terms of energy efficiency. • PID coefficients for quadrotor control are optimized with DE,PSO,CSS,GSA algorithms • The torque values to be applied to each motor were adjusted with these coefficients • The total force applied with DEPID in the z direction is smaller than others • DEPID consumes less energy than the other algorithms in the paper [ABSTRACT FROM AUTHOR]

In this study, an energy efficiency study has been carried out through the moment values of a quadrotor applied in the position control with the differential evolution algorithm (DE), which is under the evolutionary algorithms among meta-heuristic algorithms particle swarm optimization (PSO) which is among swarm-based algorithms and gravity search algorithm (GSA) among physics-based algorithms and charged system search (CSS) algorithm, which are all well-known in the literature. The efficient use of energy in electrically powered flying mechanisms is important in terms of completing its task. The aim of this study is to obtain proportional–integral–derivative (PID) coefficients that minimize the energy consumption caused by quadrotor moments. First of all, 12 PID coefficients calculated with traditional Ziegler Nicholes, and trying-error methods for the control of the quadrotor have been trained with DE, PSO, GSA and CSS algorithms and their cost values have been found for this problem. In the study, optimum PID coefficients have been found separately with each algorithm and torque data giving energy efficiency have been obtained by applying them to the state equations. Although the cost minimization convergence value obtained from DE does not look good compared to other algorithms, it has been observed that the moment values of the obtained PID coefficients in the control program are lower. In the study, it has been shown that the PID coefficients obtained from DE reach the target with approximately less energy than the maximum-moment value of the PSO, which provides the closest moment value. It can be thought that this value contributes to energy consumption in each orbital movement of the quadrotor. The simulation results show that the coefficients obtained from DE provide better energy efficiency and are superior to other algorithms compared in terms of energy efficiency. • PID coefficients for quadrotor control are optimized with DE,PSO,CSS,GSA algorithms • The torque values to be applied to each motor were adjusted with these coefficients • The total force applied with DEPID in the z direction is smaller than others • DEPID consumes less energy than the other algorithms in the paper [ABSTRACT FROM AUTHOR]

This paper presents a distributed swarm system for small fixed-wing unmanned aerial vehicles (UAVs). In particular, to perform various missions with multiple UAVs that are densely gathered and collision free, a hybrid-flocking control algorithm is synthesized by using three types of control protocols: vector field guidance (for path following/loitering), augmented Cucker–Smale (ACS) model (for collective flocking behavior), and potential field (for collision avoidance). In particular, to address the issue of conflicts between different control protocols, the adaptive ACS model is proposed and the optimization problem is formulated to determine the suitable mixing weights of control protocols. We also design the transition of multiple operation modes and communication architecture for the swarm system. The system is evaluated using the proposed hybrid-flocking control algorithm by proof-of-concept real flight experiments using 18 small fixed-wing UAVs as well as extensive numerical simulations. Flight experiments are successfully performed for multiple consecutive tasks including the individual task, circular path loitering and elliptical path loitering while avoiding collisions among UAVs. • The distributed small fixed-wing UAV swarm system is designed. • The hybrid-flocking control algorithm is proposed by combining three control protocols. • The optimal mixing weight improves the mission performance. • Numerical simulations and flight experiments validate the hybrid-flocking algorithm. [ABSTRACT FROM AUTHOR]

Multi-agent reinforcement learning (MARL) has been successfully applied in many fields. In MARL, the policy evaluation problem is one of crucial problems. In order to solve this problem, distributed Temporal-Difference (TD) learning algorithm is one of the most popular methods in a cooperative manner. Despite its empirical success, however, the theory of the adaptive variant of distributed TD learning still remain limited. To fill this gap, we propose an adaptive distributed temporal-difference algorithm (referred to as MS - ADTD) under Markovian sampling over time-varying networks. Furthermore, we rigorously analyze the convergence of MS - ADTD , the theoretical results show that the local estimation can converge linearly to the optimal neighborhood. Meanwhile, the theoretical results are verified by simulation experiments. • This paper proposes a distributed adaptive TD learning algorithm under Markovian sampling. • It theoretically analyzes the non-asymptotic convergence performance of the proposed algorithm. • It verifies the performance of the proposed algorithm by experiments. [ABSTRACT FROM AUTHOR]

Path planning is essential for robots to efficiently execute jobs in challenging settings. In this paper, we propose a new algorithm, called potential function-based sampling heuristic optimal path planning considering congestion (CCPF-RRT*), extending the standard Rapidly-exploring Random Tree star (RRT*) to address the issues of slow convergence, high path cost, and how to plan suitable paths in the presence of congested regions in the environment. First, the CCPF-RRT* algorithm adds the artificial potential field to RRT*, which can drastically cut down on the number of iterations and speed up convergence. Second, a movement cost function that takes congestion intensity and path length into account when evaluating a connection between two path nodes is created. This function is then used to direct the growth of new nodes and the updating of parent nodes so that the algorithm takes congestion intensity and path length into account when generating paths. Third, in order to reduce the path cost, the algorithm establishes parent nodes for random nodes by combining the benefits of the F-RRT* algorithm. The movement cost function is added to the process of establishing parent nodes to optimize the path cost in order to prevent the creation of parent nodes that force the path to traverse through crowded areas. In experimental simulation results, the algorithm outperforms RRT*, Q-RRT*, PQ-RRT*, and F-RRT* not only in terms of initial solution and quick convergence speed, but is also able to design an ideal path with the lowest movement cost in a crowded environment using the movement cost function. [ABSTRACT FROM AUTHOR]

Non-contact identification of cattle is the basis for cattle management, insurance claims, and traceability of milk sources and meat products. In order to lower the threshold of deep learning technology in practical application and improve the recognition accuracy, this paper uses the localization and classification paradigm, fills the image and proposes the LKA (Locate Key Area) module, which further improves the identification accuracy of cattle without using any labeling information. To avoid the impact of different collection devices, we use three brands of mobile phones and two cameras to collect data. Due to the fact that CNN features do not have scale invariance, the image is filled as a square, and the filling value is discussed. Finally, (0,0,0) is selected for filling. In order to reduce the labeling of new data, the face, torso, and body of the cattle are first detected, and then automatically trimmed to build a classification dataset. In order to further improve the accuracy of classification, LKA module is proposed. This module locates key areas based on the results of class activation map and uses shallow features for monitoring. Thereby completing fine-grained feature extraction without supervision information. Finally, the features of different parts are adjusted for fusion recognition. Experiments were conducted on the head, torso, and whole of cattle. Based on the ResNet50 network architecture, after filling the image into a square, the recognition accuracy has been improved by 0.30%, 0.21%, and 0.43%, respectively. After further adding LKA modules, the accuracy rates were improved by 0.59%, 0.59%, and 0.45%, respectively. In the fusion experiment, ResNet50_LKA and MobileNetV3_LKA achieves 100% accuracy. Experiments have realized dataset construction and fine-grained feature extraction without labeled information, and achieved high recognition accuracy, proving the effectiveness of our method. [ABSTRACT FROM AUTHOR]

High utility pattern mining (HUPM) plays a significant role in data mining technologies. Traditional HUPM algorithms may produce a large number of high utility patterns (HUPs) when the database is dense or the data is massive. To address this issue, closed high utility pattern (CHUP) mining was proposed, providing a high-level overview of the HUPs and helpful information for decision-makers. However, CHUPs do not consider the factors of period and recency. Therefore, this paper is the first to introduce period and recency into closed high utility pattern mining and proposes the CPR-Miner algorithm to mine closed periodic recent high utility patterns. These patterns have more practical value since they are closed sets of HUPs. Due to the increasing number of factors to be considered, new upper bounds and pruning strategies are also proposed, significantly improving the algorithm's efficiency. To test the performance of our algorithm and our new pruning strategies, we improved the PHM algorithm to generate the PR-Miner algorithm. Experimental results show a significant efficiency of the new pruning strategies and demonstrate that CPR-Miner outperforms the PR-Miner algorithm in all aspects. • Periodicity and recency are introduced within closed high utility pattern mining. • New data structures are designed to store the utility, period, and recency information. • Novel pruning strategies are proposed to avoid unnecessary list joint operation. • The proposed algorithm gets significant results on several datasets. [ABSTRACT FROM AUTHOR]

The high dimensionality of decision variables in large-scale multi-objective optimization problems poses significant challenges for evolutionary algorithms, which often struggle to achieve efficient search, are prone to premature convergence, and require a substantial amount of computing resources to converge to the Pareto Front. To address these challenges, a cooperative tri-population based evolutionary algorithm is proposed in this paper. Firstly, the population is partitioned into three subpopulations according to the Pareto dominant relation and crowdedness, including a subpopulation with well-distributed nondominated individuals, a subpopulation with crowded non-dominated individuals, and a subpopulation with dominated individuals. Three distinct reproduction operators are then applied to each subpopulation. The first subpopulation uses fully informed search-based reproduction to locate the true Pareto Front, while the second subpopulation adopts segment learning-based reproduction to preserve elite segments and promote exploitation. Finally, directional exploration-based reproduction is used for the third subpopulation to explore the search space and promote diversity. The proposed algorithm is capable of exploring and exploiting superior solutions through co-evolution among diverse subpopulations. Experiments are performed on 36 LSMOP benchmarks with up to 50,000 decision variables to validate the effectiveness of the proposed algorithm, which demonstrates superior performance compared to five state-of-the-art algorithms in handling large-scale multi-objective optimization problems. [ABSTRACT FROM AUTHOR]

Mobile robot path planning refers to the design of the safely collision-free path with shortest distance and least time-consuming from the starting point to the end point by a mobile robot autonomously. In this paper, a systematic review of mobile robot path planning techniques is presented. Firstly, path planning is classified into global path planning and local path planning according to the mastery of environmental information. In the global path planning, environment modeling methods and path evaluation method are introduced. The methods of environment modeling include grid method, topology method, geometric feature method and mixed representation method. In the local path planning, we introduce the sensors commonly used in the detection environment, including laser radar and visual sensor. Next, according to the characteristics of algorithms, mobile robot path planning algorithms are divided into three categories: classical algorithms, bionic algorithms and artificial intelligence algorithms. Among the classical algorithms, we introduce the cell decomposition method, sampling based method, graph search algorithm, artificial potential field method and dynamic window method. In the algorithm based on bionics, we introduce genetic algorithm, ant colony algorithm, gray wolf algorithm, etc. in detail. In artificial intelligence algorithm, we introduce neural network algorithm and fuzzy logic. Finally, we compare the key technologies of mobile robot path planning in the form of graphs and charts based on the classification statistics of the collected literature to provide references for future research. [ABSTRACT FROM AUTHOR]

Mining incrementally closed item sets is an essential problem in data mining. Once closed item sets and their frequencies are found, the frequentation of a found closed item set is always determined. In recent times, concept lattice has been used as an intermediate structure for the purpose of mining incrementally (frequent) closed sets. However, lattice operations are very time-consuming when data fluctuates as well as the memory for lattice is consumed a lot when data becomes large. This paper proposes an intermediate structure called constructive set to produce closed sets along with their occurrence frequencies. The constructive set is constructed from a set of group patterns – an extended form of bit chains. This paper also proposes algorithms based on the constructive set for mining incrementally closed sets when adding and removing transactions. The proposed algorithms along with storing and calculating on bit data have significant advantages shown through experiments and comparisons. [ABSTRACT FROM AUTHOR]