Search

Asynchronous stochastic gradient descent (ASGD) is a computationally efficient algorithm, which speeds up deep learning training and plays an important role in distributed deep learning. However, ASGD suffers from the stale gradient problem, i.e., the gradient of worker may mismatch the weight of parameter server. This problem seriously affects the model performance and even causes the divergence. To address this issue, this paper designs a dynamic adjustment scheme via the momentum algorithm, which uses both stale penalty and stale compensation , i.e., stale penalty is to reduce the trust in stale gradient, stale compensation is to compensate the hurt of stale gradient. Based on this dynamic adjustment scheme, this paper proposes a dynamic asynchronous stochastic gradient descent algorithm (DASGD), which dynamically adjusts the compensation factor and the penalty factor via stale size. Moreover, we prove that DASGD is convergent under some mild assumptions. Finally, we build a real distributed training cluster to evaluate our DASGD on Cifar10 and ImageNet datasets. Compared with four SOTA baselines, experiment results confirm the superior performance of DASGD. More specifically, our DASGD has nearly the same test accuracy as SGD on Cifar10 and ImageNet , and only uses around 27.6% and 40.8% training time that of SGD, respectively. [ABSTRACT FROM AUTHOR]

This paper proposes a fast bit-level chaotic image encryption algorithm. The goal is to make the difference between 0 and 1 in the bit-level ciphertext as small as possible. By introducing the concept of the difference between 0 and 1, constantly adjust the number of binary 0 and 1 in the ciphertext, and the difference is fed back to the encryption system until the difference reaches the preset accuracy, so as to quickly achieve a secure encryption effect. To verify the security of the algorithm, this paper selects multiple chaotic systems for testing and applies different chaotic systems to the encryption algorithm. Moreover, the proposed algorithm is compared with the bit-level and non-bit-level image encryption algorithms proposed by other scholars in recent years, and the results show that the proposed algorithm has higher security than other algorithms. [ABSTRACT FROM AUTHOR]

This paper presents a novel and robust framework for state estimation of switched finite-field networks (SFFNs), which are a class of multi-valued logical networks that can model complex systems. The paper uses the semi-tensor product (STP) method to transform the dynamical model of SFFNs into an algebraic form, which can be easily manipulated and analyzed. The paper then applies a Monte-Carlo-based sequential importance sampling (SIS) filter to estimate the state of SFFNs from noisy and incomplete observations. The paper also introduces a resampling algorithm (RA) to prevent the particle degeneracy problem (PDP), which can affect the accuracy and efficiency of the filter. Finally, the paper demonstrates the performance and advantages of the proposed multi-valued particle filter (MVPF) framework through three examples. [ABSTRACT FROM AUTHOR]

This paper proposes an interactive forest algorithm for planning single-tuned harmonic filters for electric power distribution systems. The planning of harmonic filters is a complex problem that must consider multiple objectives, including minimizing the total cost of installed filters and the individual and total harmonic distortion while adhering to operational constraints, such as the voltage drop, the current loading and the harmonic limits. Moreover, system perturbations must be considered, such as frequency variations. This paper proposes a new method for multi-objective optimization of planning single-tuned harmonic filters utilizing the proposed interactive forest algorithm with a non-differentiable multi-objective function involving continuous and discrete variables. A multi-objective optimization technique based on the proposed interactive forest algorithm is proposed. In the interactive approach, planners set expectations for each objective and use the forest algorithm to figure out the optimization problems. The results of the Pareto Front are analyzed and the answers are searched through the interactive process to identify a solution that meets the planner's expectations, allowing for practical compromise or satisfactory solutions. A case study is reported to show the superiority of the proposed method. [ABSTRACT FROM AUTHOR]

In the realm of practical problem-solving, multi-objective optimisation problems with redundant variables and indefinite objective functions (MOPRVIF) are becoming increasingly prevalent. MOPRVIF involve determining the optimal decision variables that optimise multiple objectives, leveraging the relational data of a set of variables and multiple objectives. For these problems, this paper focuses on the following two issues: one is the demand for a unified computational model to solve this problem; the other is how to improve the algorithm's deep intelligent search capability. In this regard, this paper designs a dual data-driven multi-objective optimisation method. The method used consisted of four parts: elimination of redundant variables (ERV), objective function construction (OFC), selection evolution operator (SEO), and multi-objective evolutionary algorithm (MOEA). MOEA was the main focus of the method. ERV is data preparation and variable selection according to multiple objectives. OFC involves constructing the relationship model between variables and objectives, and a high-accuracy model is important for guaranteeing reliable results. Furthermore, SEO can adjust the evolution operator during a deep search. This is an important guarantee for deep, intelligent search. MOEA combined OFC and SEO to form the final solution algorithm—Dual Data Driven Multi-Objective Evolutionary Algorithm (DDMOEA). DDMOEA was explored using two different disciplinary problems of drug compound optimisation and wild blueberry cultivation and benchmarks were selected. The first two problem domains are distinct. The first problem is more complex than the second; however, both encompass redundant variables and indefinite objective functions. Benchmarks are utilised independently to gauge the profound intelligent search capability. The experiments affirm that the dual data-driven optimization approach proposed in this paper is effective, practical, and scalable. • Provide a unified solution for optimisation problems with redundant variables and indefinite objective functions. • Propose a data-driven multi-objective optimisation algorithm. • Designing an adaptive optimiser to enhance the algorithm's depth search capability. [ABSTRACT FROM AUTHOR]

Semi-supervised graph clustering with constraints has received considerable attention in the last decade. They use pre-given constraints to guide the clustering process and improve the performance. Nonetheless, most of related research works still face two main issues: (1) cannot-link problem: how to ensure that instances under the cannot-link constraint are located in different clusters as much as possible. (2) The high computational cost caused by eigendecomposition. To tackle these problems, this paper proposes a novel method for directly solving the clustering indicator matrix without high complexity computational steps. Then, we propose an effective and simple algorithm guided by the pre-given constraints, which is able to simultaneously optimize the associated multiple pairs of constraints, allowing the associated constraints to achieve the related co-optimal solution. Extensive experiments are conducted to verify that our approach significantly reduces the violation rate of the prior constraints and the clustering performance of the proposed method outperforms seven other state-of-the-art semi-supervised clustering methods on 16 real benchmark datasets. [ABSTRACT FROM AUTHOR]

The task of mining high-utility itemsets in a database given a minimum threshold is attracting more and more interest due to its many applications. Existing algorithms such as the vertical ones have the advantages of high scalability, efficiency and extensibility. However, they depend on a costly join operation to generate new itemsets. To overcome this limitation, this paper proposes a novel vertical algorithm, FOTH (Fast sOrted iTemset searcH), which employs a novel effective data structure, the IndexSet. The IndexSet self-propagates to produce sub-IndexSets, eliminating the need to perform join operations, which considerably reduces the memory and computation requirements. Experiments were conducted on eight benchmark databases to compare the performance of FOTH with four state-of-the-art list-based algorithms. The results show that FOTH outperforms the other algorithms on dense databases. [ABSTRACT FROM AUTHOR]

Incremental feature selection methods have gained increasing research attention as they improve the efficiency of feature selection for dynamic datasets. Multigranulation rough set, as an extension of rough set theory, allows for a comprehensive and rational analysis of problems from multiple hierarchical and granular perspectives. However, existing research on granularity partitioning relies on the decision maker's subjective experience, which lacks convincing power. In this paper, we propose a generalized multigranulation neighborhood rough set based on weight partition model, using a matrix form. We discuss several properties and define a new entropy measure to evaluate feature importance. A heuristic feature selection algorithm is developed based on this entropy to search for the optimal subset. Furthermore, we discuss dynamic updating mechanism and design two incremental feature selection algorithms. Finally, we conduct experiments on 12 public datasets to evaluate the performance of the proposed algorithms and validate their effectiveness and efficiency in feature selection for both static and dynamic datasets. [ABSTRACT FROM AUTHOR]

Dynamic multi-objective optimization is a relatively challenging problem within the field of multi-objective optimization. Nevertheless, these problems have significant real-world applications. The key to addressing dynamic multi-objective problems effectively is promptly tracking changes in the Pareto set (PS) and Pareto front (PF). Dynamic multi-objective optimization encompasses various types of problems, and a single-type response strategy proves effective for some specific scenarios. However, as problem complexity and diversity increase, a single-type response strategy often falls short in solving dynamic multi-objective optimization problems. To address this issue, this paper proposes a hybrid response dynamic multi-objective optimization algorithm. The suggested algorithm utilizes the multi-arm bandit model to adaptively adjust the proportion of different response strategies for each type of multi-objective optimization problem. Furthermore, it achieves rapid convergence through an enhanced two-stage MOEA/D. Experiments demonstrate the effectiveness of the strategies employed in the proposed algorithm and its competitiveness compared to other state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Partial label learning aims to learn from training instances, each of which is associated with a set of candidate labels but only one is a ground-truth label. Feature selection is an effective method to improve the generalization capability of the learning model; however, partial label feature selection work is exceptionally challenging due to the limitation and ambiguity of label information. Therefore, this paper proposes a partial label feature selection algorithm based on label disambiguation and neighborhood mutual information. Firstly, neighborhood granularity is utilized to determine the neighborhoods of instances to disambiguate the candidate labels. Secondly, based on label confidence induced by disambiguation, feature relevance and redundancy are measured by neighborhood mutual information, which avoids the negative impact of data discretization on feature selection and directly handles continuous features. Concurrently, the kappa coefficient is employed to estimate the label consistency for describing the influences of feature changes on the label space. Then, the significance of each feature is evaluated by fusing feature relevance, feature redundancy, and label consistency. Finally, the effectiveness of the proposed algorithm is verified by comparing the proposed algorithm with four base classifiers and other feature selection methods. Furthermore, the feasibility of the proposed disambiguation method is demonstrated through comparison with four state-of-the-art disambiguation methods. [ABSTRACT FROM AUTHOR]

Feature selection addresses the dimensionality-reduction problem by determining a subset of available features, which facilitates the construction of effective prediction models. However, with regression tasks where the target variable is continuous and imbalanced, ordinary feature-selection techniques cannot be simply used without adjustment. This paper proposes SerEnet, a novel method of sparse feature selection in imbalanced regression, which explores both feature selection and estimation simultaneously by minimizing the Squared Error-Relevance with respect to a cutoff t (S E R t), subject to a sparse penalty. Specifically, S E R t considers the performance of target variables with relevance greater than t in the target-variable domain and emphasizes the error of rare values. Moreover, SerEnet can effectively identify features that contribute significantly to rare cases, thereby reducing the dominant influence of common instances on feature selection, and improving prediction performance for both rare values and overall data. Experimental results on simulated and real datasets show that SerEnet outperformed several algorithms in terms of prediction performance on continuous-imbalanced data. [ABSTRACT FROM AUTHOR]

Characterization of the optimization problem is a crucial task in many recent optimization research topics (e.g., explainable algorithm performance assessment, and automated algorithm selection and configuration). The state-of-the-art approaches use exploratory landscape analysis to represent the optimization problem, where for each one, a set of features is extracted using a set of candidate solutions sampled by a sampling strategy over the whole decision space. This paper proposes a novel representation of continuous optimization problems by encoding the information found in the interaction between an algorithm and an optimization problem. The new problem representation is learned using the information from the states/positions in the optimization run trajectory (i.e., the candidate solutions visited by the algorithm). With the novel representation, the problem can be characterized dynamically during the optimization run, instead of using a set of candidate solutions from the whole decision space that have never been observed by the algorithm. The novel optimization problem representation is called Opt2Vec and uses an autoencoder type of neural network to encode the information found in the interaction between an optimization algorithm and optimization problem into an embedded subspace. The Opt2Vec representation efficiency is shown by enabling different optimization problems to be successfully identified using only the information obtained from the optimization run trajectory. • Representation learning is applied on each individual population of optimization process trajectory. • The Opt2Vec representations are captured through the algorithm's behavior (its populations). • The Opt2Vec representations are suitable for dynamic problem characterization. • The Opt2Vec representations are invariant to simple transformations (shifting/scaling). • The Opt2Vec representations are scalable over different problem dimensions. [ABSTRACT FROM AUTHOR]

Differential Privacy (DP) is applied in Federated Learning (FL) for defending against various privacy attacks. Existing methods based on Gaussian mechanism require the operations of clipping and adding noise, leading to significant accuracy degradation. In this paper, we propose a novel FL scheme named DPFL-LMG to provide user-level DP guarantee while maintaining a high model accuracy. Our main idea is to mitigate the negative effects of the clipping on the model convergence by decreasing the L 2 norm of local updates and the cross-client update variance. Specifically, our method includes two techniques, Local Momentum Updates (LMU) and Gradients Filtering (GF). LMU combines local updates of different rounds in a momentum way. It can significantly decrease the cross-client update variance by weakening the gradient noise in local updates caused by stochastic gradient descent (SGD) algorithm. GF estimates the gradient noise in each element of local updates by observing the element-wise variance. Elements with large noise are considered unnecessary and are zeroed out for the reduction of local update norms. We theoretically analyze the privacy guarantee and the convergence of our method. Experiments demonstrate that DPFL-LMG can effectively mitigate the accuracy degradation caused by clipping and outperform previous DPFL methods in the accuracy. • Providing rigorous user-level DP guarantee, while maintaining a high accuracy and a fast convergence speed. • Mitigating the negative effects of the clipping on the model convergence. • Decreasing the L 2 norm of local updates and cross-client update variance. • A theoretical analysis of the privacy guarantee and the model convergence. [ABSTRACT FROM AUTHOR]

Due to their good approximation accuracy and local fitting characteristics, the moving least squares (MLS) and moving total least squares (MTLS) methods are widely used in various engineering fields. However, neither of these two methods is robust and they cannot effectively deal with outliers in measurement data. To eliminate the negative influence of outliers and achieve robust reconstruction, a novel MTLS method is proposed in this paper, which introduces local Bayesian estimation combined with clustering using representatives (CURE) algorithm. In the support domain, this method adopts a two-step process to remove the abnormal points and adjust the weights of discrete points through compound weighting. Bayesian estimation is first performed on discrete points to derive the reference model, and the residuals are calculated as the input of CURE clustering. The points with large residuals are classified into one cluster and removed. The remaining points undergo repeated processing until the iteration concludes. A gradient weight function based on the residuals and a compact support weight function are combined to determine the final estimated value using weighted Bayesian estimation. The simulations and experiments demonstrate that the proposed reconstruction method achieves excellent accuracy and robustness, surpassing several existing methods when handling highly contaminated datasets. [ABSTRACT FROM AUTHOR]

• An approach is proposed for estimating monotonic change point in multistage process. • The quality of the process at each stage is characterized by simple linear profiles. • Two SVM algorithms are proposed for identifying the occurrence and type of change. • MLE is proposed for estimating both single and multiple monotonic change points. • The performance of the proposed approach is evaluated under various scenarios. In this paper, a hybrid method is proposed for estimating monotonic change points in multistage processes where, at each stage, the quality of the process is represented by simple linear profile models. In the case of monotonic changes, the change type is not known a priori, and the only assumption is that the changes are non-decreasing or non-increasing in nature. In the proposed method, a support vector machine (SVM) algorithm is first developed to identify the parameters experiencing a change in each process stage. Then, the second SVM algorithm is applied to identify the type of change occurring in the related parameters. Finally, considering the identified parameters and change types, the maximum likelihood estimator (MLE) is proposed to estimate the change points of the process under different scenarios, including single and multiple change points, upward and downward step changes, increasing and decreasing linear trends, shift in both parameters of the profiles, shift in different stages of the process, and weak and strong autocorrelation coefficients. The performance of the proposed method is evaluated through extensive simulation experiments. The results indicate that the proposed method recognizes the patterns and estimates the monotonic change points accurately. [ABSTRACT FROM AUTHOR]

In multi-label data, the importance of each label within the logical label vector varies for each sample, and there exist inherent correlations among the labels. However, the logical label vector fails to capture these nuances. Consequently, relying solely on this vector for feature selection in multi-label data results in underutilization of supervisory information. To address this issue, this paper introduces a novel label enhancement algorithm. This algorithm leverages neighborhood information derived from features to transform the logical label vector into a label distribution that effectively reflects label differences and correlations. Subsequently, we propose a feature selection algorithm tailored for multi-label data, which incorporates both the transformed label distribution and mutual information. This algorithm not only accounts for the mutual information between features and label distributions but also captures the mutual information among features themselves. Finally, we evaluate our proposed feature selection algorithm against five state-of-the-art multi-label feature selection algorithms on ten publicly available datasets. The experimental results reveal that our algorithm outperforms its competitors in six distinct evaluation metrics, achieving an average performance improvement of approximately 9%. This substantial enhancement underscores the efficacy of our algorithm in handling complex multi-label data. [ABSTRACT FROM AUTHOR]

Sparse recovery under nonnegativity and sum-to-one constraints is a special form of the linear regression problem, where the solution is required to simultaneously satisfy sparsity, nonnegativity, and sum-to-one restraints. Existing algorithms for this task mainly utilize the penalty technique to convert the sparsity constraint into a regularization term. Therefore, the sparsity is determined via tuning the associated penalty parameter, which is time-consuming in practice. This paper exploits projected gradient descent to directly tackle the constrained problem without involving the penalty parameter and ℓ 0 -norm approximation. The addition of the ℓ 0 -norm constraint with a specific upper bound enables the proposed algorithm to explicitly control sparsity. The developed method is termed as modified iterative hard thresholding (MIHT), comprised of two iterative steps, namely, gradient descent and nonconvex projection. For the latter, the constraint set consists of the ℓ 0 -norm, nonnegativity, and sum-to-one constraints. We devise an efficient algorithm to address the nonconvex projection and then prove that this method produces an optimal solution. Furthermore, we establish the convergence of the MIHT, including objective value and variable sequence. Numerical experiments using financial and hyperspectral data demonstrate that the MIHT is superior to state-of-the-art methods in terms of prediction error and recovery accuracy. [ABSTRACT FROM AUTHOR]

In large-scale distributed learning, the direct application of traditional inference is often not feasible, because it may contain multiple themes, such as communication costs, privacy issues, and Byzantine failures. Nowadays, the internet is vulnerable to attacks, and Byzantine failures frequently occur. For copying with Byzantine failures, the paper develops two Byzantine-robust distributed learning algorithms under a framework of communication-efficient surrogate likelihood. In our algorithms, we adopt the δ -approximate compressors, including sign-based operator and top k sparsification, to improve communication efficiency, and an unsophisticated thresholding of local gradient norms to guard against Byzantine failures. For accelerating convergence and achieving an optimal statistical error rate, error feedback is exploited in the second algorithm. The two algorithms are robust to arbitrary adversaries, although Byzantine workers don't adhere to the mandated compression mechanism. We explicitly establish statistical error rates, which imply that our algorithms don't sacrifice the quality of learning, and attain the order-optimal under some settings. In addition, we provide a trade-off between compression and adversary in the presence of Byzantine worker machines. Extensive numerical experiments validate our theoretical results and demonstrate a good performance of our algorithms. [ABSTRACT FROM AUTHOR]

Frequent itemset mining (FIM) is a widely used task that extracts frequently occurring itemsets from data. Plenty of deterministic algorithms are available for this daunting task. However, experimental studies have not considered that data heterogeneity significantly impacts the algorithms' performance, giving rise to unfair comparisons and biased conclusions. This paper seeks to advance by comparing cutting-edge algorithms using various frequency thresholds, considering the resulting data heterogeneity. An extensive experimental study is carried out, including the number of itemsets mined per second as the performance quality measure to compare algorithms. The experiments include defining eight metrics to quantify data heterogeneity, and their values vary the algorithms' performance. The results revealed that some techniques (hypercube decomposition and k -items machine) are essential to achieve excellent performance on any dataset, and most algorithms behave similarly well when they include those techniques. As a final important point, different threshold values produce dissimilar data subsets (data heterogeneity is not an immutable data characteristic), so a previous study on the database characteristics with a few minimum support thresholds could be beneficial to select the best-suited FIM algorithm beforehand. [ABSTRACT FROM AUTHOR]

In multi-label learning, label enhancement (LE) aims to recover label distributions from logical labels, thereby reinforcing supervision information in the training set. Existing LE algorithms mainly leverage pairwise similarities between instances to recover label distributions. However, symmetric similarities, which are widely used, are incapable of reflecting individual differences between instances. In this paper, we propose an asymmetric influence between instances to fully explore the information contained in logical labels. First, we build a bipartite network and employ the mass diffusion algorithm on it to calculate the asymmetric influence between instances. Such asymmetry distinguishes the influence of an instance on others from the influence it receives from others. Second, to cope with the new influence, we construct a graph with self-connections. In this way, we allow instances to be influenced by themselves. Extensive experiments on thirteen benchmark datasets demonstrate the superiority of the proposed method over six state-of-the-art LE algorithms. • We exploit asymmetric influence between instances when label enhancement. • This asymmetry distinguishes individual differences to better exploit the correlation between instances. • Results validate its effectiveness in comparison with several state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Multi-objective evolutionary algorithms (MOEAs) have been successfully applied in recommender systems, where the recommendation is modeled as a multi-objective optimization problem by considering both accuracy and non-accuracy metrics. However, to recommend to all users, the population individuals in most of the exiting MOEA-based recommendation algorithms are usually encoded in the form of matrix encoding, and with the increased number of users and items, it will lead to a large search space. To this end, in this paper, we propose a novel evolutionary multitasking algorithm MOREM to tackle the challenge of multi-objective recommendations, where a related auxiliary task is constructed by both user and item reduction to help solving the complex original task by knowledge transfer. Specifically, a task generation strategy for creating auxiliary task is firstly proposed to cluster users by user-rating information (i.e., user reduction), where only a part of important items (i.e., item reduction) are recommended for the central user within each cluster with the aim to greatly reduce the size of the matrix encoding. In addition, a novel knowledge transfer mechanism is proposed, which can effectively achieve knowledge migration between the two tasks. Experimental results on real-world datasets show that MOREM outperforms several state-of-the-art algorithms for multi-objective recommendations. [ABSTRACT FROM AUTHOR]

In the context of selecting features for multi-scale interval valued decision table (M I V D T), conventional research approaches encounter difficulties including elevated data complexity, diminished computational efficiency, and limited model generalization capacity. To overcome these difficulties, feature selection methods based on contradictory state sequence (CSS) and fuzzy contradictory state sequence (F C S S) are proposed in this paper. Initially, M I V D T is established. According to the characteristics of interval value, a more thorough and impartial metric for assessing the inclusion degree is suggested, along with similarity relation based on the metric. Subsequently, the introduction of the first contradictory object allows for the delineation of contradictory state and fuzzy contradictory state, which serve to characterize the consistency of M I V D T. These two concepts are proposed to enhance the accuracy and efficiency of capturing key information in intricate data. Additionally, rapid feature selection algorithms are suggested, which rely on CSS and F C S S. In contrast to conventional feature selection approaches, the algorithms introduced in this study demonstrate superior computational efficiency and enhanced generalization capabilities when confronted with intricate datasets. In twelve open source datasets, the upper limit for the quantity of objects is 110341, and the average accuracy values of experiments under two parameter combinations are 99.54% and 97.08% respectively. The results of experiments demonstrate that the algorithms introduced are capable of efficiently identifying a novel feature subset from intricate datasets within a shorter timeframe. [ABSTRACT FROM AUTHOR]

Multimodal multi-objective optimization problem (MMOP) is a variant of the multi-objective problem (MOP), which has multiple optimal Pareto sets (PSs) mapping to the same or similar Pareto front (PF) from the decision space to the objective space. As algorithms must obtain multiple Pareto Sets, addressing MMOPs presents greater challenges for algorithms. To address the issue, the paper proposes a novel multimodal multi-objective differential evolution algorithm named MMODE_SDNR. Firstly, the algorithm sorts the population using the fast non-dominated sorting method and the Shortest Distance (SD) values. SD represents the shortest Euclidean distance between an individual and its nearest neighboring individual, calculated jointly in the decision space and the objective space. Secondly, a definition of the similarity between an individual and its neighbors in both spaces is provided, based on which a novel environmental selection strategy for the algorithm, called the nearest Neighbor-Repulsion (nNR) is devised. The strategy removes the individuals that have similarity with one another to maximize the retention of different individuals. Moreover, it keeps the diversity of PSs in the decision space and retains more solutions that exhibit a similar or same PF in the objective space. Finally, the proposed MMODE_SDNR has undergone several tests on the benchmark functions from CEC 2019 and CEC 2020 to assess its performance. The experimental results indicate that MMODE_SDNR excels in three metrics, outperforming other several state-of-the-art MMO algorithms. Its strong convergence and focus on decision space quality ensure competitiveness. [ABSTRACT FROM AUTHOR]

Solving nonlinear equation systems (NESs) is one of the most important tasks in numerical computing. The NESs usually have multiple roots, and quickly locating their roots in a single algorithm run with a limited number of iterations has always been the algorithm improvement direction. In order to further enhance the efficiency of the existing methods, a density clustering-based differential evolution algorithm (DCDE) for the NESs problem solving is proposed in this paper. Firstly, density-based spatial clustering of applications with noise (DBSCAN) is used to divide the population into clusters and noises, which provides a better direction for population evolution. Secondly, a cluster evaluation factor is proposed, which not only divides the clusters into excellent clusters and non-excellent clusters, but also prevents the migration of excellent clusters and maintain the diversity of populations. Then, a migration strategy and individual generation mechanism are proposed to guide non-excellent clusters to migrate to promising regions. Finally, combined with an archive technique, the individuals which satisfy the solution requirements are archived and randomly initialized to further improve population diversity. To verify the effectiveness of the proposed algorithm, comparative experimental results of the propsoed DCDE with other state-of-the-art algorithms for thirty NESs problems solving show that the proposed DCDE is the most effective algorithm to locate more roots in a single run. [ABSTRACT FROM AUTHOR]

Clustering, as an essential technique in unsupervised learning, plays a pivotal role in the fields of data mining and machine learning. However, the classic K -means clustering algorithm has intrinsic drawbacks such as sensitivity to initial cluster centers, susceptibility to a local optimal solution, and challenges in handling data uncertainty. To address these problems, this paper proposes an artificial hummingbird algorithm (AHA)-based three-way K -means clustering algorithm, called AHA-3WKM. First, AHA is introduced to address the problems of sensitivity to initial cluster centers and local optima. Second, a fitness function of AHA is specifically constructed to find the best initial clustering centers so that the hummingbirds can search for high-quality food sources, i.e., the global optimum cluster centers. Third, a three-way clustering approach is utilized to capture information about data uncertainty. In this way, the results of clustering are divided into three distinct regions based on the relationship between objects and clusters. The experimental results demonstrate that AHA-3WKM has good performance, and enhances the stability and the accuracy of clustering results. • AHA is introduced to address the problems of the sensitivity to initial cluster centers and the proneness to local optima. Hummingbirds are treated as data points, which dynamically update their strategies and effectively find the optimal cluster centers during multiple iterations. • A fitness function is designed based on the clustering principle of "birds of a feather flock together", with the aim of simplifying calculations, which enhances the specificity and practicality of K-means algorithm. • An AHA-based three-way K-means clustering algorithm (i.e., AHA-3WKM) is proposed. The clustering process is initialized with cluster centers optimized by AHA, and the results are represented in three regions, which can capture the uncertainty within the datasets. [ABSTRACT FROM AUTHOR]

Clustering analysis is a powerful tool for discovering potential knowledge in datasets. However, numerous existing clustering algorithms suffer from heavy reliance on parameter settings and cannot cluster complex manifold data very well. Moreover, although non-parameter clustering algorithms aim to lower the usage threshold, their actual performance in clustering is often unsatisfactory. Addressing how to achieve similar clustering effects for numerous data points with only a few core data points during clustering is a valuable consideration. To alleviate these challenges, a non-parameter clustering algorithm is proposed and named N on- p arameter C lustering Algorithm based on C hain P ropagation and N atural Neighbor (NPCCPN) in this paper, by jointly using chain propagation and natural neighbor. Specifically, NPCCPN identifies core points through chain propagation and clusters them using the saturated neighborhood graph. This makes the core data extraction and clustering process efficient and non-parameter. Finally, the performance of the method is validated on 15 complex synthetic datasets and 10 real datasets from public UCI database. The experimental results show that the effectiveness and superiority of the proposed algorithm. Purity scores first. ACC scores second without parameters, but the score first algorithm needs to set the parameters. [ABSTRACT FROM AUTHOR]

• An improved gaining-sharing knowledge-based algorithm named MSGSK is proposed. • A parameter adjustment strategy is developed to adjust relative parameters of MSGSK. • A backtracking differential mutation strategy is designed to enrich the population diversity. • A strategy selection mechanism is presented to integrate the former two strategies. • MSGSK extracts more accurate parameters for five PV models. The determination of photovoltaic (PV) model parameters has essential theoretical and practical significance for the performance evaluation, power monitoring, and power generation efficiency calculation of PV systems. In this paper, a multi-strategy gaining-sharing knowledge-based algorithm (MSGSK) is developed to determine these parameters. In our previous work, it has been demonstrated that gaining-sharing knowledge-based algorithm (GSK) is well suited for solving the concerned problem. To enhance its performance, a parameter adjustment strategy is developed to adjust the knowledge rate and knowledge ratio of GSK. Besides, a backtracking differential mutation strategy by combining the mutation scheme of differential evolution and the updating scheme of backtracking search optimization algorithm is developed to enrich the population diversity. Furthermore, a strategy selection mechanism is introduced to integrate the former two strategies to balance exploration and exploitation in different stages of the evolutionary process. The suggested MSGSK algorithm is applied to five PV cases (SDM, DDM, Photowatt-PW201, STM6-40/36, and STP6-120/36). From the experimental data, it can be observed that MSGSK extracts the PV model parameters more precisely than the basic GSK. Furthermore, it exhibits faster convergence speed and higher accuracy compared to other advanced algorithms found in the literature. [ABSTRACT FROM AUTHOR]

• Analyze DE algorithm two key issues: parameter sensitivity and search imbalance. • A complementary mutation strategy based on complementary symmetric selection and a Gaussian distribution mutation is proposed. • Data fusion-based parameter adaptation method proposed for parameter rationalization. • A novel variant of DE called DACDE is designed and results shown competitive edge. As an excellent optimization algorithm widely used to solve various practical problems, differential evolution (DE) algorithm has few parameters, yet its performance is significantly affected by these parameters. To address this issue, this paper presents a novel variant of DE called DACDE, which utilizes data fusion-based parameter adaptation and a complementary mutation strategy. Most parameter adaptation methods based on successful history only analyze the mean of parameters and ignore their degree of dispersion. In DACDE, the successful parameter distribution is recorded and described by both the mean and variance. Data fusion is then used to combine records and generate an estimated distribution, which is applied to a Gaussian distribution to generate new parameters. Inspired by opposition-based learning, we introduce a complementary mutation strategy. This strategy employs a symmetric selection mechanism to adapt to the varying search abilities required by the algorithm at different stages. The new variant is verified on 32 single-objective functions from CEC 2011 and 2014 benchmark suites, and the results show that DACDE is competitive compared to other 29 evolutionary algorithms. [ABSTRACT FROM AUTHOR]

In a modern distributed learning framework, the speeds of intra-worker calculation and inter-worker communication may differ by 1,000 times. It is advisable to perform expensive calculations on worker machines and communicate as few rounds as possible. In this paper, we propose three novel distributed sparse learning algorithms (EDSL-ET, EDSL-IBCD, and EDSL-IBCD-ET) in high dimensions, which are efficient in both computation and communication. Algorithm EDSL-ET utilizes the Top k sparse technique to reduce communication costs and adopts error feedback to guarantee convergence. Algorithm EDSL-IBCD greatly reduces communication costs by introducing an independent block coordinate descent method. Algorithm EDSL-IBCD-ET takes advantage of EDSL-ET and EDSL-IBCD algorithms for extra-high dimensional feature learning, which is the best one whether in communication, computation, and stability. We give theoretical guarantees of the three algorithms under mild conditions, which match the performance of the centralized algorithm. Extensive experiments on simulated and real data validate our theoretical analysis and demonstrate that the proposed algorithms perform well with just a few rounds of communications. [ABSTRACT FROM AUTHOR]

Federated learning (FL) enables the full utilization of decentralized training without raw data. However, various attacks still threaten the training process of FL. To address these concerns, differential privacy (DP) and secure multi-party computation (SMC) are applied, but these methods may result in low accuracy and heavy training load. Moreover, the high communication consumption of FL in resource-constrained devices is also a challenging problem. In this paper, we propose a novel SMC algorithm for the FL (FL-IPFE) to protect the local gradients. It does not require a trusted third party (TTP) and is more suitable for FL. Furthermore, we propose a secure and efficient FL algorithm (SEFL), which applies compressed sensing (CS) and all-or-nothing transform (AONT) to minimize the number of transmitted and encrypted model updates. Additionally, our FL-IPFE is used to encrypt the last element of the preprocessed parameters for guaranteeing the security of the entire local model updates. Meanwhile, the issue of participant dropouts is also taken into account. Theoretical analyses demonstrate that our proposed algorithms can aggregate model updates with high security. Finally, experimental evaluation reveals that our SEFL possesses higher efficiency compared to other state-of-the-art works, while providing comparable model accuracy and strong privacy guarantees. [ABSTRACT FROM AUTHOR]

Mining co-location patterns hidden in spatial data is crucial for spatial association discovery, and it has broad prospects in many applications. H igh U tility C o-location P attern M ining (HUCPM) further takes the utility factor of spatial features into consideration, so it is more realistic compared with the traditional co-location pattern mining. However, HUCPM is more difficult, since the Apriori-like pruning technique does not apply. To address this problem, we firstly suggest two novel pruning strategies to trim the pattern search space. Then, a series of optimizing techniques are presented to speed up the pattern utility ratio calculation of each candidate. Based on above techniques, a fast HUCPM algorithm is proposed, which searches for high utility co-locations involved in each pattern branch via a depth-extending manner and equips with a heuristic strategy to enhance the effect of pruning techniques. Moreover, we theoretically prove the completeness and correctness of the proposed algorithm, and discuss its algorithmic complexity. On multiple spatial datasets, we conduct substantial experiments to reveal the superiority of our algorithm in efficiency and scalability, as well as the effectiveness of the proposed technique. Particularly, the proposed algorithm in this paper runs faster than other baselines for several times to several orders of magnitude. [ABSTRACT FROM AUTHOR]

Multi-view subspace clustering aims to leverage the consensus and complementary information embedded within the data and effectively fuse the multiple views. However, existing methods have limitations in fully exploring the complementary information. Tensor-based methods tend to overlook the underlying structures, while latent representation-based methods struggle to capture high-order correlations, both of which are important components of complementary information. In this paper, we introduce a novel method called complete multi-view subspace clustering (CMVSC) using visible and latent views. The latent view with the underlying structures is obtained by the latent representation, along with visible views (i.e., original features) constitute the complete multi-view data. Then, the low-rank tensor constraint is imposed on the complete views to comprehensively explore the consensus and complementary information. Besides, to effectively combine the advantages of visible and latent views, we devise an auto-weighted strategy that can automatically assign the ideal weights to each view. Finally, an efficient alternating iterative algorithm using the augmented Lagrange multiplier method is designed to solve the CMVSC model. Experimental results show that our model outperforms the state-of-the-art counterparts while achieving almost perfect clustering performance on several benchmark datasets. [ABSTRACT FROM AUTHOR]

In this paper, a new differential evolution (DE) algorithm is presented for solving global optimization problems by guiding the individuals adaptively to explore different decision regions and setting the control parameters properly for individuals. To consolidate algorithm efficiency and preserve population diversity, we first introduce a proration-based mutation strategy. This strategy dynamically selects promising individuals from population as guiders and proportionately allocates the individuals following them based on their fitness values and unsuccessful updates. Besides, we design a multi-segment mixed parameter setting to provide suitable parameters for individuals, considering feedback information from the population, individual requirements, and parameters' interaction. Additionally, we devise an adaptive population maintaining mechanism to refresh the population by replacing individuals with higher unsuccessful updates and removing those with poor fitness values. Unlike previous DE versions, our new algorithm dynamically and appropriately explores promising areas, refines parameter suitability, and accelerates convergence by updating and removing unpromising individuals. This enhances the algorithm search efficiency and achieves a good balance between exploration and exploitation. Finally, we evaluate the proposed algorithm's performance by comparing it with 17 typical or state-of-the-art algorithms on IEEE CEC2014 and CEC2017 test suites. Experimental results indicate that the proposed algorithm is a more promising optimizer. [ABSTRACT FROM AUTHOR]

Clustering methods are being applied to a wider range of scenarios involving more complex datasets, where the shapes of clusters tend to be arbitrary. In this paper, we propose a novel Path-based Valley-seeking clustering algorithm for arbitrarily shaped clusters. This work aims to seek the valleys among clusters and then individually extract clusters. Three vital techniques are used in this algorithm. First, path distance (minmax distance) is employed to transform the irregular boundaries among clusters, that is density valleys, into perfect spherical shells. Second, a suitable density measurement, k -distance, is employed to make adjustment on Minimum Spanning Tree, by which a robust minmax distance is calculated. Third, we seek the transformed density valleys by determining their centers and radii. Based on the vital techniques, the main contributions of the proposed algorithm can be summarized as follows. First, the clusters are wrapped in spherical shells after the distance transformation, making the extraction process efficient even with clusters of arbitrary shape. Second, adjusted Minimum Spanning Tree enhances the robustness of minmax distance under different kinds of noise. Last, the number of clusters does not need to be inputted or decided manually due to the individual extraction process. After applying the proposed algorithm to several commonly used synthetic datasets, the results indicate that the Path-based Valley-seeking algorithm is accurate and efficient. The algorithm is based on the dissimilarity of objects, so it can be applied to a wide range of fields. Its performance on real-world datasets illustrates its versatility. [ABSTRACT FROM AUTHOR]

Constrained multi-objective optimization problems (CMOPs) require a delicate balance between satisfying constraints and optimizing objectives. Existing constrained multi-objective evolutionary algorithms (CMOEAs) often struggle to balance convergence, diversity, and feasibility, especially when dealing with CMOPs that have complex feasible regions. This paper proposes a multi-task-based self-organizing mapping evolutionary algorithm (MTSOM) to tackle this challenge, which includes a main and auxiliary task. Two populations independently optimize two tasks without considering constraints in the early stage. Subsequently, in the middle stage, both tasks explore the distribution structure of the population in parallel by employing a novel constraint-to-constraint self-organizing mapping (SOM) approach. In the late stage, the main task fully considers feasibility, while the auxiliary task focuses solely on the highest priority constraints. This approach enables rapid convergence toward feasible regions. To evaluate MTSOM's effectiveness, we conducted a series of experiments on five benchmark suites. Results indicate that MTSOM is competitive when compared to other state-of-the-art CMOEAs. Additionally, our proposed constraint-to-constraint SOM is superior in handling complex CMOPs. [ABSTRACT FROM AUTHOR]

Large-volume and high-dimensional big datasets are being generated quickly. They are expected to provide data-driven solutions for various pressing challenges such as cybersecurity, credit card fraud, network intrusion detection, etc. The visual assessment of clustering tendency (VAT) family of algorithms provides an excellent tool for assessing clustering tendency and subsequent clustering of these novel datasets to extract groups and anomalies to better understand the data generation phenomenon. However, VAT and improved VAT (iVAT) algorithms are plagued with O (n 2) time and space complexity as they use Prim's algorithm for building Euclidean minimum spanning tree (EMST) of the data points, leaving it inapplicable to large datasets. This paper develops three novel time- and memory-scalable algorithms for fast computation of VAT EMST by reducing the search space of the following possible EMST edge and using an efficient data structure, the k-d tree. Next, we develop a novel approach to compute iVAT reordered dissimilarity image (RDI) in a novel memory-preserving manner using the EMST edge lengths calculated previously without having to compute memory intensive n × n VAT RDI. We experimented on several synthetic and real-life datasets to showcase that the proposed approaches can help quickly assess their clustering tendency in a memory-saving manner. We also demonstrate the applicability of the proposed methods for the anomaly detection task in large volumes of high-dimensional data. [ABSTRACT FROM AUTHOR]

Algorithm selection technology, aimed at choosing the most suitable algorithm for a given machine learning task, has achieved significant success in the domain of 2D vision. However, few studies have explored its application to the 3D point cloud domain. The rapid proliferation and development of 3D point cloud classification algorithms underscore the urgent need for exploration in selecting these algorithms. In this paper, we propose a novel meta-learning-based 3D classification approach, termed 3D meta-classification , to address this gap. The approach operates at both the base-level and meta-level phases. At the base level, candidate 3D classification algorithms constantly classify various 3D datasets, recording their classification performance as empirical knowledge. As for the meta-level phase, it specifically tailors the 3D meta-knowledge generator, 3D meta-feature extractor, and 3D meta-database constructor to establish the 3D meta-database, capturing the relationship between the 3D meta-features and empirical knowledge. Leveraging the 3D meta-database, the 3D classification meta-learner trains the meta-model and predicts suitable algorithms for new incoming 3D datasets. Extensive experiments are conducted to select the best-performing classification algorithm for specific 3D datasets from ModelNet40. The results demonstrate the effectiveness of the proposed 3D meta-classification model; the accuracies of one-from-two and one-from-three algorithm selection tasks reach over 98% and 80%, respectively. [ABSTRACT FROM AUTHOR]

As a population-based stochastic optimization algorithm, Gravitational Search Algorithm (GSA) has attracted numerous interests and has been applied in various applications. However, GSA has drawbacks such as uneven search and premature convergence in practical applications. This paper specifically explains the inherent characteristic of GSA in prioritizing the center position. Correspondingly, an improvement strategy of fitness normalization with mass shift is proposed, creating a situation where gravity and repulsion are mixed. Then, the global best mechanism with weights is incorporated into the particle's velocity update formula, which compensates for the difficulties in the later exploitation stage. Finally, an empirical formula for the initial gravitational constant related to the size of the solution space is proposed, which enhances the global search ability together with the former strategy. 12 shifted benchmark functions are used to construct 20 optimization problems ranging from 2 to 120 dimensions. The average performance of the proposed algorithm, other GSA and well-known algorithms are compared under the same budget. The results demonstrate that the proposed GSA not only effectively addresses the drawbacks of GSA and maintains good performance, but also exhibits strong competitiveness compared to various similar algorithms. [ABSTRACT FROM AUTHOR]

Traditional nonprofiling side-channel analysis frequently adopts divide-and-conquer strategy to recover the secret key of a cryptographic algorithm. Only a single key byte is used, whereas the remaining bytes are considered extraneous noise. If the cryptographic algorithm is implemented in hardware with parallel processing, the strategy will lead to the inefficient use of information. The combination of intelligent algorithms and side-channel analysis offers a different idea for nonprofiling methods for parallel hardware implementations. It transforms the problem of key recovery into key optimization. In this paper, we adopt this idea and propose an efficient heuristic framework to assist correlation power analysis. The multipoint hill-climbing algorithm will be used to find the optimal key candidates with higher scores, and the correct key will be found in the optimal key candidates by key enumeration. Multiple bytes of the secret key are used to take full advantage of the helpful information. Besides, two search strategies are introduced to overcome the efficiency problem, and two key enumeration strategies are introduced to solve the key recovery problem when the traces are insufficient. Experimental results on the public DPA Contest V1 dataset show that the heuristic framework performs better than other classical methods, verifying its effectiveness. [ABSTRACT FROM AUTHOR]

The application of semantic technologies, particularly ontologies, in the realm of multi-objective evolutionary algorithms is overlook despite their effectiveness in knowledge representation. In this paper, we introduce MOODY, an ontology specifically tailored to formalize these kinds of algorithms, encompassing their respective parameters, and multi-objective optimization problems based on a characterization of their search space landscapes. MOODY is designed to be particularly applicable in automatic algorithm configuration, which involves the search of the parameters of an optimization algorithm to optimize its performance. In this context, we observe a notable absence of standardized components, parameters, and related considerations, such as problem characteristics and algorithm configurations. This lack of standardization introduces difficulties in the selection of valid component combinations and in the re-use of algorithmic configurations between different algorithm implementations. MOODY offers a means to infuse semantic annotations into the configurations found by automatic tools, enabling efficient querying of the results and seamless integration across diverse sources through their incorporation into a knowledge graph. We validate our proposal by presenting four case studies. • A semantic model is proposed for consolidating multi-objective optimization knowledge. • A knowledge graph is presented, as well as the tools to include new RDF data. • The model allows for semantic reasoning on evolutionary algorithm configurations. • The proposal is validated on three use cases with data from experiments. • Moody enables knowledge-based recommendations of configurations for algorithms. [ABSTRACT FROM AUTHOR]

Despite the huge success of reinforcement learning (RL) in solving many difficult problems, its Achilles heel has always been sample inefficiency. On the other hand, in RL, taking advantage of prior knowledge, intentionally or unintentionally, has usually been avoided, so that, training an agent from scratch is common. This not only causes sample inefficiency but also endangers safety –especially during exploration. In this paper, we help the agent learn from the environment by using the pre-existing (but not necessarily exact or complete) solution for a task. Our proposed method can be integrated with any RL algorithm developed based on policy gradient and actor-critic methods. The results on five tasks with different difficulty levels by using two well-known actor-critic-based methods as the backbone of our proposed method (SAC and TD3) show our success in greatly improving sample efficiency and final performance. We have gained these results alongside robustness to noisy environments at the cost of just a slight computational overhead, which is negligible. [ABSTRACT FROM AUTHOR]

This paper introduces a novel algorithm for solving the point-to-point shortest path problem in a static regular 8-neighbor connectivity (G8) grid. This algorithm can be seen as a generalization of Hadlock algorithm to G8 grids, and is shown to be theoretically equivalent to the relaxed A ⁎ (R A ⁎) algorithm in terms of the provided solution's path length, but with substantial time and memory savings, due to a completely different computation strategy, based on defining a set of lookup matrices. Through an experimental study on grid maps of various types and sizes (1290 runs on 43 maps), it is proven to be 2.25 times faster than R A ⁎ and 17 times faster than the original A ⁎ , in average. Moreover, it is more memory-efficient, since it does not need to store a G score matrix. [ABSTRACT FROM AUTHOR]

As one of the main research tasks in software testing, automated test case generation based on path coverage (ATCG-PC) aims to achieve maximum path coverage with a minimized set of test cases. In ATCG-PC, the correlation among the dimensions of test cases is widely utilized in academia to minimize the search efforts of the search-based algorithm. Nevertheless, the information related to target path selection is not utilized, which leads to blind decision-making by the search-based algorithm during the target path selection. Therefore, this paper proposes an enhanced scatter search strategy by using opposition-based learning. An arithmetic optimization algorithm is also proposed to solve ATCG-PC based on the enhanced scatter search strategy, namely, ESSENT. The ESSENT algorithm selects the path with the lowest path entropy among the uncovered paths as the target path, and generates new test cases that cover the target path by modifying the dimensions of the existing test cases. The performance of the ESSENT algorithm is evaluated on six iFogSim subprograms and six Stanford coreNLP subprograms. Experiment results show that the ESSENT algorithm achieves a higher convergence rate than other state-of-the-art algorithms. Furthermore, it enables maximum path coverage with fewer test cases. [ABSTRACT FROM AUTHOR]

The joint order batching and picker routing problem (JOBPRP) is a combinatorial optimization problem that occurs during in the order-picking operation of warehouse management. It consists of simultaneous assignment of customer orders to batches and routing of picking operations in the batches such that the total travel distance is minimized. In this paper, we develop a memetic algorithm with fuzzy-based population control (MA-FPC) to solve the JOBPRP. MA-FPC takes advantage of the hierarchical nature of the problem to design a batch exchange crossover (BEX) operator and a two-level local improvement procedure. BEX transfers complete batches from parent solutions to offsprings. Local improvement systematically employs a combined heuristic and exact routing approach to solve the picker routing problem. Furthermore, fuzzy-based population control is used to dynamically regulate population diversity during the search, which removes individuals from or adds individuals to the population. In numerical experiments, MA-FPC significantly outperforms state-of-the-art point-based and population-based algorithms developed for the order batching problem and JOBPRP with respect to solution quality in reasonable running times. Our algorithm improves the previous best-known solutions for 57 of the 64 benchmark instances. [ABSTRACT FROM AUTHOR]

A variety of algorithms has been proposed to mitigate the overestimation bias of Q-learning. These algorithms reduce the estimation of maximum Q-value, i.e., homogeneous update. As a result, some of these algorithms such as Double Q-learning suffer from the underestimation bias. Different from these algorithms, this paper proposes a heterogeneous update idea. It aims to enlarge the normalized gap between Q-value corresponding to the optimal action and that corresponding to the other actions. Based on heterogeneous update , we design HetUp Q-learning. More specifically, HetUp Q-learning increases the normalized gap by overestimating Q-value corresponding to the optimal action and underestimating Q-value corresponding to the other actions. However, one limitation is that our HetUp Q-learning takes the optimal action as input to decide whether a state-action pair should be overestimated or underestimated. To address this challenge, we apply a softmax strategy to estimate the optimal action and obtain HetUpSoft Q-learning. We also extend HetUpSoft Q-learning to HetUpSoft DQN for high-dimensional environments. Extensive experiment results show that our proposed methods outperform SOTA baselines drastically in different settings. In particular, HetUpSoft DQN improves the average score per episode over SOTA baselines by at least 55.49% and 32.26% in the Pixelcopter and Breakout environments, respectively. [ABSTRACT FROM AUTHOR]

Balancing multiple objectives and various constraints is crucial for effectively solving constrained multi-objective optimization problems (CMOPs). Excessive focus on either convergence or feasibility may not result in favorable outcomes of the algorithm. To confront this challenge, this paper proposes a cooperative evolutionary algorithm named SaE-CMO, which aims to achieve a harmonious balance between convergence and feasibility by extracting valuable information from both feasible and infeasible regions. To achieve this, SaE-CMO employs a dual-population approach to enhance search progress, consisting of a main population, Population1, and an auxiliary population, Population2. These two populations complement each other to achieve optimal results. A newly proposed self-adaptive epsilon method is employed in both Population1 and Population2, using different comparison criteria to select next population from mating pools, respectively. Population2 can retain some solutions that are well-constrained but poorly converged, thereby preserving information about both the constrained and the unconstrained Pareto front. This property enables Population2 to assist Population1 in maintaining diversity in certain complex CMOPs. To verify the effectiveness of SaE-CMO, we conduct experiments on three benchmark test instances and four real-world CMOPs with some related state-of-the-art constrained multi-objective optimization algorithms, experimental results prove that the proposed algorithm outperforms the compared algorithms. [ABSTRACT FROM AUTHOR]

Differential privacy (DP), as a promising privacy-preserving model, has attracted great interest from researchers in recent years. At present, research on the combination of deep learning and DP is active. In contrast, another widely used artificial intelligence technique, swarm intelligence (SI), has received little attention in the context of DP even though it also triggers privacy concerns. For this reason, this paper attempts to combine DP and SI for the first time and proposes a general differentially private swarm intelligence algorithm framework (DPSIAF). By utilizing the exponential mechanism, this framework can easily develop existing SI algorithms into the private versions. As examples, we apply it to four popular SI algorithms, and the related analyses demonstrate the effectiveness of our DPSIAF. More importantly, our private algorithms also exhibit some interesting experimental phenomena. In some cases, their performance is not strictly affected by the privacy budget, and one of them even outperforms its non-private version. These findings are different from the conventional cognition, which indicates DP owns the potential to serve as an optimization tool for some SI algorithms. Our study may provide a new perspective on DP and promote the synergy between metaheuristic optimization community and privacy computing community. [ABSTRACT FROM AUTHOR]

Border-Peeling algorithm is a recently proposed density based clustering algorithm. The method of peeling off border points by continuous iteration and calculating the density influence value by using the Gaussian kernel distance makes the algorithm more complex. At the same time, there is a risk of excessive peeling of small clusters on unbalanced datasets, which leads to a large number of noise misidentification. In order to reduce the time consumption and improve the clustering performance on unbalanced datasets, this paper proposes a non-iterative border-peeling clustering algorithm. First, the potential core points are determined by the centroids of k-nearest neighbor. Secondly, the points with lower local relative density in the core points and the points with higher relative density in the border points are exchanged to complete the distinction between the core points and the border points. Then basic DBSCAN method is used to cluster core points and noise points. Finally, the associations between the border points and the core points are based on the number of reverse nearest neighbors of the border points in the core points. Our method has achieved competitive results on 10 synthetic datasets and 8 UCI real-world datasets. [ABSTRACT FROM AUTHOR]

The time series database is a specialized type of database specifically designed for storing and analyzing time series data. Compression of time series data is crucial for its performance. However, efficiently compressing time series data, particularly floating-point data, remains challenging. Existing compression algorithms are efficient for only a limited range of data patterns, indicating a lack of self-adaptation. In this paper, we propose an effective and A daptive lossless F loating-point C ompression algorithm AFC for time series databases. We devise four unique compression strategies, and based on the data patterns, AFC dynamically selects the appropriate strategy. These strategies handle data compression for diverse data patterns, enhancing the compression ratio and efficiency. The most suitable strategy is employed to achieve an optimal compression ratio. We compared our AFC algorithm with four state-of-the-art compression algorithms, namely Gorilla, FPC, TSXor, and Chimp, as well as various general-purpose compression algorithms such as LZ4 and Snappy. Experimental results demonstrate that our algorithm achieves an improvement of at least 20% in compression ratio and even up to 100% on certain datasets. [ABSTRACT FROM AUTHOR]

Optimal granularity selection is a key issue in rough set, by which decision rules can be generated to assign corresponding decision labels for new samples. The current evaluation criteria of optimal granularity mainly focus on the experimental effect of prediction algorithms, thus ignoring the theoretical relationship between granularity and the generalization ability of algorithms. In this paper, we study the issue overlooked above and propose a novel framework for optimal granularity selection with theoretical guarantee. In our framework, a rough set-based prediction algorithm that incorporates a global confidence as scoring function is introduced, which is characterized by the granularity-based loss function. On the bias, the relationship between granularity and the generalization ability of algorithm is studied by introducing the stability theory in machine learning. The generalization error bound of algorithm is derived as a theoretical guarantee for optimal granularity selection. Finally, a novel optimal granularity selection strategy with theoretical guarantee is proposed, which is combined with existing attribute reduction methods to design the attribute re-reduction algorithms for generate optimal granularity by reducing the relatively unimportant attributes. Numerical experiments verify the rationality of optimal granularity selection framework and prove the effectiveness of the optimal granularity generated by the attribute re-reduction algorithms. [ABSTRACT FROM AUTHOR]