Search

This document is a special issue of the journal Algorithms, featuring selected papers from the journal's editorial board members from 2022 and 2023. The issue includes 16 research papers covering a range of topics such as game theory, fault detection in cellular networks, optimization algorithms, machine learning, cryptocurrency trading, and more. Each paper presents its own unique research findings and methodologies. The issue aims to showcase the diverse research interests and expertise of the journal's editorial board members. [Extracted from the article]

Fake news is an explosive subject, being undoubtedly among the most controversial and difficult challenges facing society in the present-day environment of technology and information, which greatly affects the individuals who are vulnerable and easily influenced, shaping their decisions, actions, and even beliefs. In the course of discussing the gravity and dissemination of the fake news phenomenon, this article aims to clarify the distinctions between fake news, misinformation, and disinformation, along with conducting a thorough analysis of the most widely read academic papers that have tackled the topic of fake news research using various machine learning techniques. Utilizing specific keywords for dataset extraction from Clarivate Analytics' Web of Science Core Collection, the bibliometric analysis spans six years, offering valuable insights aimed at identifying key trends, methodologies, and notable strategies within this multidisciplinary field. The analysis encompasses the examination of prolific authors, prominent journals, collaborative efforts, prior publications, covered subjects, keywords, bigrams, trigrams, theme maps, co-occurrence networks, and various other relevant topics. One noteworthy aspect related to the extracted dataset is the remarkable growth rate observed in association with the analyzed subject, indicating an impressive increase of 179.31%. The growth rate value, coupled with the relatively short timeframe, further emphasizes the research community's keen interest in this subject. In light of these findings, the paper draws attention to key contributions and gaps in the existing literature, providing researchers and decision-makers innovative viewpoints and perspectives on the ongoing battle against the spread of fake news in the age of information. [ABSTRACT FROM AUTHOR]

In the context of academic expert finding, this paper investigates and compares the performance of information retrieval (IR) and machine learning (ML) methods, including deep learning, to approach the problem of identifying academic figures who are experts in different domains when a potential user requests their expertise. IR-based methods construct multifaceted textual profiles for each expert by clustering information from their scientific publications. Several methods fully tailored for this problem are presented in this paper. In contrast, ML-based methods treat expert finding as a classification task, training automatic text classifiers using publications authored by experts. By comparing these approaches, we contribute to a deeper understanding of academic-expert-finding techniques and their applicability in knowledge discovery. These methods are tested with two large datasets from the biomedical field: PMSC-UGR and CORD-19. The results show how IR techniques were, in general, more robust with both datasets and more suitable than the ML-based ones, with some exceptions showing good performance. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) algorithms are playing a crucial role in transforming healthcare by enhancing the quality, accessibility, and efficiency of medical care, research, and operations. These algorithms enable healthcare providers to offer more accurate diagnoses, predict outcomes, and customize treatments to individual patient needs. AI also improves operational efficiency by automating routine tasks and optimizing resource management. However, there are challenges to adopting AI in healthcare, such as data privacy concerns and potential biases in algorithms. Collaboration among stakeholders is necessary to ensure ethical use of AI and its positive impact on the field. AI also has applications in medical research, preventive medicine, and public health. It is important to recognize that AI should augment, not replace, the expertise and compassionate care provided by healthcare professionals. The ethical implications and societal impact of AI in healthcare must be carefully considered, guided by fairness, transparency, and accountability principles. Several research papers in this special issue explore the application of AI algorithms in various aspects of healthcare, such as gait analysis for Parkinson's disease diagnosis, human activity recognition, heart disease prediction, compliance assessment with clinical protocols, epidemic management, neurological complications identification, fall prevention, leukemia diagnosis, and genetic clinical pathways. These studies demonstrate the potential of AI in improving medical diagnostics, patient monitoring, and personalized care. [Extracted from the article]

Cardiovascular disease is the leading cause of global mortality and responsible for millions of deaths annually. The mortality rate and overall consequences of cardiac disease can be reduced with early disease detection. However, conventional diagnostic methods encounter various challenges, including delayed treatment and misdiagnoses, which can impede the course of treatment and raise healthcare costs. The application of artificial intelligence (AI) techniques, especially machine learning (ML) algorithms, offers a promising pathway to address these challenges. This paper emphasizes the central role of machine learning in cardiac health and focuses on precise cardiovascular disease prediction. In particular, this paper is driven by the urgent need to fully utilize the potential of machine learning to enhance cardiovascular disease prediction. In light of the continued progress in machine learning and the growing public health implications of cardiovascular disease, this paper aims to offer a comprehensive analysis of the topic. This review paper encompasses a wide range of topics, including the types of cardiovascular disease, the significance of machine learning, feature selection, the evaluation of machine learning models, data collection & preprocessing, evaluation metrics for cardiovascular disease prediction, and the recent trends & suggestion for future works. In addition, this paper offers a holistic view of machine learning's role in cardiovascular disease prediction and public health. We believe that our comprehensive review will contribute significantly to the existing body of knowledge in this essential area. [ABSTRACT FROM AUTHOR]

The interconnection of Operational Technology (OT) and Information Technology (IT) has created new opportunities for remote management, data storage in the cloud, real-time data transfer over long distances, or integration between different OT and IT networks. OT networks require increased attention due to the convergence of IT and OT, mainly due to the increased risk of cyber-attacks targeting these networks. This paper focuses on the analysis of different methods and data processing for protocol recognition and traffic classification in the context of OT specifics. Therefore, this paper summarizes the methods used to classify network traffic, analyzes the methods used to recognize and identify the protocol used in the industrial network, and describes machine learning methods to recognize industrial protocols. The output of this work is a comparative analysis of approaches specifically for protocol recognition and traffic classification in OT networks. In addition, publicly available datasets are compared in relation to their applicability for industrial protocol recognition. Research challenges are also identified, highlighting the lack of relevant datasets and defining directions for further research in the area of protocol recognition and classification in OT environments. [ABSTRACT FROM AUTHOR]

In decentralized systems, the quest for heightened security and integrity within blockchain networks becomes an issue. This survey investigates anomaly detection techniques in blockchain ecosystems through the lens of unsupervised learning, delving into the intricacies and going through the complex tapestry of abnormal behaviors by examining avant-garde algorithms to discern deviations from normal patterns. By seamlessly blending technological acumen with a discerning gaze, this survey offers a perspective on the symbiotic relationship between unsupervised learning and anomaly detection by reviewing this problem with a categorization of algorithms that are applied to a variety of problems in this field. We propose that the use of unsupervised algorithms in blockchain anomaly detection should be viewed not only as an implementation procedure but also as an integration procedure, where the merits of these algorithms can effectively be combined in ways determined by the problem at hand. In that sense, the main contribution of this paper is a thorough study of the interplay between various unsupervised learning algorithms and how this can be used in facing malicious activities and behaviors within public and private blockchain networks. The result is the definition of three categories, the characteristics of which are recognized in terms of the way the respective integration takes place. When implementing unsupervised learning, the structure of the data plays a pivotal role. Therefore, this paper also provides an in-depth presentation of the data structures commonly used in unsupervised learning-based blockchain anomaly detection. The above analysis is encircled by a presentation of the typical anomalies that have occurred so far along with a description of the general machine learning frameworks developed to deal with them. Finally, the paper spotlights challenges and directions that can serve as a comprehensive compendium for future research efforts. [ABSTRACT FROM AUTHOR]

Forecasting the generation of solar power plants (SPPs) requires taking into account meteorological parameters that influence the difference between the solar irradiance at the top of the atmosphere calculated with high accuracy and the solar irradiance at the tilted plane of the solar panel on the Earth's surface. One of the key factors is cloudiness, which can be presented not only as a percentage of the sky area covered by clouds but also many additional parameters, such as the type of clouds, the distribution of clouds across atmospheric layers, and their height. The use of machine learning algorithms to forecast the generation of solar power plants requires retrospective data over a long period and formalising the features; however, retrospective data with detailed information about cloudiness are normally recorded in the natural language format. This paper proposes an algorithm for processing such records to convert them into a binary feature vector. Experiments conducted on data from a real solar power plant showed that this algorithm increases the accuracy of short-term solar irradiance forecasts by 5–15%, depending on the quality metric used. At the same time, adding features makes the model less transparent to the user, which is a significant drawback from the point of view of explainable artificial intelligence. Therefore, the paper uses an additive explanation algorithm based on the Shapley vector to interpret the model's output. It is shown that this approach allows the machine learning model to explain why it generates a particular forecast, which will provide a greater level of trust in intelligent information systems in the power industry. [ABSTRACT FROM AUTHOR]

This paper presents various machine learning methods with different text features that are explored and evaluated to determine the authorship of the texts in the example of the Azerbaijani language. We consider techniques like artificial neural network, convolutional neural network, random forest, and support vector machine. These techniques are used with different text features like word length, sentence length, combined word length and sentence length, n-grams, and word frequencies. The models were trained and tested on the works of many famous Azerbaijani writers. The results of computer experiments obtained by utilizing a comparison of various techniques and text features were analyzed. The cases where the usage of text features allowed better results were determined. [ABSTRACT FROM AUTHOR]

Deep neural networks (DNNs) have gained prominence in various applications, but remain vulnerable to adversarial attacks that manipulate data to mislead a DNN. This paper aims to challenge the efficacy and transferability of two contemporary defense mechanisms against adversarial attacks: (a) robust training and (b) adversarial training. The former suggests that training a DNN on a data set consisting solely of robust features should produce a model resistant to adversarial attacks. The latter creates an adversarially trained model that learns to minimise an expected training loss over a distribution of bounded adversarial perturbations. We reveal a significant lack in the transferability of these defense mechanisms and provide insight into the potential dangers posed by L ∞ -norm attacks previously underestimated by the research community. Such conclusions are based on extensive experiments involving (1) different model architectures, (2) the use of canonical correlation analysis, (3) visual and quantitative analysis of the neural network's latent representations, (4) an analysis of networks' decision boundaries and (5) the use of equivalence of L 2 and L ∞ perturbation norm theories. [ABSTRACT FROM AUTHOR]

The rapid expansion and pervasive reach of the internet in recent years have raised concerns about evolving and adaptable online threats, particularly with the extensive integration of Machine Learning (ML) systems into our daily routines. These systems are increasingly becoming targets of malicious attacks that seek to distort their functionality through the concept of poisoning. Such attacks aim to warp the intended operations of these services, deviating them from their true purpose. Poisoning renders systems susceptible to unauthorized access, enabling illicit users to masquerade as legitimate ones, compromising the integrity of smart technology-based systems like Network Intrusion Detection Systems (NIDSs). Therefore, it is necessary to continue working on studying the resilience of deep learning network systems while there are poisoning attacks, specifically interfering with the integrity of data conveyed over networks. This paper explores the resilience of deep learning (DL)—based NIDSs against untethered white-box attacks. More specifically, it introduces a designed poisoning attack technique geared especially for deep learning by adding various amounts of altered instances into training datasets at diverse rates and then investigating the attack's influence on model performance. We observe that increasing injection rates (from 1% to 50%) and random amplified distribution have slightly affected the overall performance of the system, which is represented by accuracy (0.93) at the end of the experiments. However, the rest of the results related to the other measures, such as PPV (0.082), FPR (0.29), and MSE (0.67), indicate that the data manipulation poisoning attacks impact the deep learning model. These findings shed light on the vulnerability of DL-based NIDS under poisoning attacks, emphasizing the significance of securing such systems against these sophisticated threats, for which defense techniques should be considered. Our analysis, supported by experimental results, shows that the generated poisoned data have significantly impacted the model performance and are hard to be detected. [ABSTRACT FROM AUTHOR]

The Linear Hinges Model (LHM) is an efficient approach to flexible and robust one-dimensional curve fitting under stringent high-noise conditions. However, it was initially designed to run in a single-core processor, accessing the whole input dataset. The surge in data volumes, coupled with the increase in parallel hardware architectures and specialised frameworks, has led to a growth in interest and a need for new algorithms able to deal with large-scale datasets and techniques to adapt traditional machine learning algorithms to this new paradigm. This paper presents several ensemble alternatives, based on model selection and combination, that allow for obtaining a continuous piecewise linear regression model from large-scale datasets using the learning algorithm of the LHM. Our empirical tests have proved that model combination outperforms model selection and that these methods can provide better results in terms of bias, variance, and execution time than the original algorithm executed over the entire dataset. [ABSTRACT FROM AUTHOR]

The paper is devoted to the theoretical and numerical analysis of the two-step method, constructed as a modification of Polyak's heavy ball method with the inclusion of an additional momentum parameter. For the quadratic case, the convergence conditions are obtained with the use of the first Lyapunov method. For the non-quadratic case, sufficiently smooth strongly convex functions are obtained, and these conditions guarantee local convergence.An approach to finding optimal parameter values based on the solution of a constrained optimization problem is proposed. The effect of an additional parameter on the convergence rate is analyzed. With the use of an ordinary differential equation, equivalent to the method, the damping effect of this parameter on the oscillations, which is typical for the non-monotonic convergence of the heavy ball method, is demonstrated. In different numerical examples for non-quadratic convex and non-convex test functions and machine learning problems (regularized smoothed elastic net regression, logistic regression, and recurrent neural network training), the positive influence of an additional parameter value on the convergence process is demonstrated. [ABSTRACT FROM AUTHOR]

When dealing with engineering design problems, designers often encounter nonlinear and nonconvex features, multiple objectives, coupled decision making, and various levels of fidelity of sub-systems. To realize the design with limited computational resources, problems with the features above need to be linearized and then solved using solution algorithms for linear programming. The adaptive linear programming (ALP) algorithm is an extension of the Sequential Linear Programming algorithm where a nonlinear compromise decision support problem (cDSP) is iteratively linearized, and the resulting linear programming problem is solved with satisficing solutions returned. The reduced move coefficient (RMC) is used to define how far away from the boundary the next linearization is to be performed, and currently, it is determined based on a heuristic. The choice of RMC significantly affects the efficacy of the linearization process and, hence, the rapidity of finding the solution. In this paper, we propose a rule-based parameter-learning procedure to vary the RMC at each iteration, thereby significantly increasing the speed of determining the ultimate solution. To demonstrate the efficacy of the ALP algorithm with parameter learning (ALPPL), we use an industry-inspired problem, namely, the integrated design of a hot-rolling process chain for the production of a steel rod. Using the proposed ALPPL, we can incorporate domain expertise to identify the most relevant criteria to evaluate the performance of the linearization algorithm, quantify the criteria as evaluation indices, and tune the RMC to return the solutions that fall into the most desired range of each evaluation index. Compared with the old ALP algorithm using the golden section search to update the RMC, the ALPPL improves the algorithm by identifying the RMC values with better linearization performance without adding computational complexity. The insensitive region of the RMC is better explored using the ALPPL—the ALP only explores the insensitive region twice, whereas the ALPPL explores four times throughout the iterations. With ALPPL, we have a more comprehensive definition of linearization performance—given multiple design scenarios, using evaluation indices (EIs) including the statistics of deviations, the numbers of binding (active) constraints and bounds, the numbers of accumulated linear constraints, and the number of iterations. The desired range of evaluation indices (DEI) is also learned during the iterations. The RMC value that brings the most EIs into the DEI is returned as the best RMC, which ensures a balance between the accuracy of the linearization and the robustness of the solutions. For our test problem, the hot-rolling process chain, the ALP returns the best RMC in twelve iterations considering only the deviation as the linearization performance index, whereas the ALPPL returns the best RMC in fourteen iterations considering multiple EIs. The complexity of both the ALP and the ALPPL is O(n2). The parameter-learning steps can be customized to improve the parameter determination of other algorithms. [ABSTRACT FROM AUTHOR]

Reinforcement learning is an important technique in various fields, particularly in automated machine learning for reinforcement learning (AutoRL). The integration of transfer learning (TL) with AutoRL in combinatorial optimization is an area that requires further research. This paper employs both AutoRL and TL to effectively tackle combinatorial optimization challenges, specifically the asymmetric traveling salesman problem (ATSP) and the sequential ordering problem (SOP). A statistical analysis was conducted to assess the impact of TL on the aforementioned problems. Furthermore, the Auto_TL_RL algorithm was introduced as a novel contribution, combining the AutoRL and TL methodologies. Empirical findings strongly support the effectiveness of this integration, resulting in solutions that were significantly more efficient than conventional techniques, with an 85.7% improvement in the preliminary analysis results. Additionally, the computational time was reduced in 13 instances (i.e., in 92.8% of the simulated problems). The TL-integrated model outperformed the optimal benchmarks, demonstrating its superior convergence. The Auto_TL_RL algorithm design allows for smooth transitions between the ATSP and SOP domains. In a comprehensive evaluation, Auto_TL_RL significantly outperformed traditional methodologies in 78% of the instances analyzed. [ABSTRACT FROM AUTHOR]

This paper reviews the application of artificial neural network (ANN) models to time series prediction tasks. We begin by briefly introducing some basic concepts and terms related to time series analysis, and by outlining some of the most popular ANN architectures considered in the literature for time series forecasting purposes: feedforward neural networks, radial basis function networks, recurrent neural networks, and self-organizing maps. We analyze the strengths and weaknesses of these architectures in the context of time series modeling. We then summarize some recent time series ANN modeling applications found in the literature, focusing mainly on the previously outlined architectures. In our opinion, these summarized techniques constitute a representative sample of the research and development efforts made in this field. We aim to provide the general reader with a good perspective on how ANNs have been employed for time series modeling and forecasting tasks. Finally, we comment on possible new research directions in this area. [ABSTRACT FROM AUTHOR]

The rapid proliferation of new technologies such as Internet of Things (IoT), cloud computing, virtualization, and smart devices has led to a massive annual production of over 400 zettabytes of network traffic data. As a result, it is crucial for companies to implement robust cybersecurity measures to safeguard sensitive data from intrusion, which can lead to significant financial losses. Existing intrusion detection systems (IDS) require further enhancements to reduce false positives as well as enhance overall accuracy. To minimize security risks, data analytics and machine learning can be utilized to create data-driven recommendations and decisions based on the input data. This study focuses on developing machine learning models that can identify cyber-attacks and enhance IDS system performance. This paper employed logistic regression, support vector machine, decision tree, and random forest algorithms on the UNSW-NB15 network traffic dataset, utilizing in-depth exploratory data analysis, and feature selection using correlation analysis and random sampling to compare model accuracy and effectiveness. The performance and confusion matrix results indicate that the Random Forest model is the best option for identifying cyber-attacks, with a remarkable F1 score of 97.80%, accuracy of 98.63%, and low false alarm rate of 1.36%, and thus should be considered to improve IDS system security. [ABSTRACT FROM AUTHOR]

This document is a special issue of the journal "Algorithms" focused on algorithms in data classification. It provides an overview of different subcategories of data classification, such as binary classification, multi-class classification, and text classification. The issue includes twelve papers that cover a wide range of applications, including class imbalance in Gaussian mixture models, blood cell classification using deep learning techniques, student dropout prediction in online education, and energy consumption in industrial plants. Other topics covered include the use of artificial intelligence models for rehabilitation guidance, hyperparameter optimization of artificial neural networks, feature selection in big data classification, and the classification of acute psychological stress and physical activity using wristband devices. The issue also includes specialized software for data classification and the use of artificial intelligence models for intelligent wear monitoring. [Extracted from the article]

This study investigates the problem of decentralized dynamic resource allocation optimization for ad-hoc network communication with the support of reconfigurable intelligent surfaces (RIS), leveraging a reinforcement learning framework. In the present context of cellular networks, device-to-device (D2D) communication stands out as a promising technique to enhance the spectrum efficiency. Simultaneously, RIS have gained considerable attention due to their ability to enhance the quality of dynamic wireless networks by maximizing the spectrum efficiency without increasing the power consumption. However, prevalent centralized D2D transmission schemes require global information, leading to a significant signaling overhead. Conversely, existing distributed schemes, while avoiding the need for global information, often demand frequent information exchange among D2D users, falling short of achieving global optimization. This paper introduces a framework comprising an outer loop and inner loop. In the outer loop, decentralized dynamic resource allocation optimization has been developed for self-organizing network communication aided by RIS. This is accomplished through the application of a multi-player multi-armed bandit approach, completing strategies for RIS and resource block selection. Notably, these strategies operate without requiring signal interaction during execution. Meanwhile, in the inner loop, the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm has been adopted for cooperative learning with neural networks (NNs) to obtain optimal transmit power control and RIS phase shift control for multiple users, with a specified RIS and resource block selection policy from the outer loop. Through the utilization of optimization theory, distributed optimal resource allocation can be attained as the outer and inner reinforcement learning algorithms converge over time. Finally, a series of numerical simulations are presented to validate and illustrate the effectiveness of the proposed scheme. [ABSTRACT FROM AUTHOR]

In recent years, with the rapid advancements in Natural Language Processing (NLP) technologies, large models have become widespread. Traditional reinforcement learning algorithms have also started experimenting with language models to optimize training. However, they still fundamentally rely on the Markov Decision Process (MDP) for reinforcement learning, and do not fully exploit the advantages of language models for dealing with long sequences of problems. The Decision Transformer (DT) introduced in 2021 is the initial effort to completely transform the reinforcement learning problem into a challenge within the NLP domain. It attempts to use text generation techniques to create reinforcement learning trajectories, addressing the issue of finding optimal trajectories. However, the article places the training trajectory data of reinforcement learning directly into a basic language model for training. Its aim is to predict the entire trajectory, encompassing state and reward information. This approach deviates from the reinforcement learning training objective of finding the optimal action. Furthermore, it generates redundant information in the output, impacting the final training effectiveness of the agent. This paper proposes a more reasonable network model structure, the Action-Translator Transformer (ATT), to predict only the next action of the agent. This makes the language model more interpretable for the reinforcement learning problem. We test our model in simulated gaming scenarios and compare it with current mainstream methods in the offline reinforcement learning field. Based on the presented experimental results, our model demonstrates superior performance. We hope that introducing this model will inspire new ideas and solutions for combining language models and reinforcement learning, providing fresh perspectives for offline reinforcement learning research. [ABSTRACT FROM AUTHOR]

Ambiguous optical illusions have been a paradigmatic object of fascination, research and inspiration in arts, psychology and video games. However, accurate computational models of perception of ambiguous figures have been elusive. In this paper, we design and train a deep neural network model to simulate human perception of the Necker cube, an ambiguous drawing with several alternating possible interpretations. Defining the weights of the neural network connection using a quantum generator of truly random numbers, in agreement with the emerging concepts of quantum artificial intelligence and quantum cognition, we reveal that the actual perceptual state of the Necker cube is a qubit-like superposition of the two fundamental perceptual states predicted by classical theories. Our results finds applications in video games and virtual reality systems employed for training of astronauts and operators of unmanned aerial vehicles. They are also useful for researchers working in the fields of machine learning and vision, psychology of perception and quantum–mechanical models of human mind and decision making. [ABSTRACT FROM AUTHOR]

Bayesian networks (BNs) are a foundational model in machine learning and causal inference. Their graphical structure can handle high-dimensional problems, divide them into a sparse collection of smaller ones, underlies Judea Pearl's causality, and determines their explainability and interpretability. Despite their popularity, there are almost no resources in the literature on how to compute Shannon's entropy and the Kullback–Leibler (KL) divergence for BNs under their most common distributional assumptions. In this paper, we provide computationally efficient algorithms for both by leveraging BNs' graphical structure, and we illustrate them with a complete set of numerical examples. In the process, we show it is possible to reduce the computational complexity of KL from cubic to quadratic for Gaussian BNs. [ABSTRACT FROM AUTHOR]

Artificial immune systems (AIS), as nature-inspired algorithms, have been developed to solve various types of problems, ranging from machine learning to optimization. This paper proposes a novel hybrid model of AIS that incorporates cellular automata (CA), known as the cellular automata-based artificial immune system (CaAIS), specifically designed for dynamic optimization problems where the environment changes over time. In the proposed model, antibodies, representing nominal solutions, are distributed across a cellular grid that corresponds to the search space. These antibodies generate hyper-mutation clones at different times by interacting with neighboring cells in parallel, thereby producing different solutions. Through local interactions between neighboring cells, near-best parameters and near-optimal solutions are propagated throughout the search space. Iteratively, in each cell and in parallel, the most effective antibodies are retained as memory. In contrast, weak antibodies are removed and replaced with new antibodies until stopping criteria are met. The CaAIS combines cellular automata computational power with AIS optimization capability. To evaluate the CaAIS performance, several experiments have been conducted on the Moving Peaks Benchmark. These experiments consider different configurations such as neighborhood size and re-randomization of antibodies. The simulation results statistically demonstrate the superiority of the CaAIS over other artificial immune system algorithms in most cases, particularly in dynamic environments. [ABSTRACT FROM AUTHOR]

In the context of TinyML, many research efforts have been devoted to designing forward topologies to support On-Device Learning. Reaching this target would bring numerous advantages, including reductions in latency and computational complexity, stronger privacy, data safety and robustness to adversarial attacks, higher resilience against concept drift, etc. However, On-Device Learning on resource constrained devices poses severe limitations to computational power and memory. Therefore, deploying Neural Networks on tiny devices appears to be prohibitive, since their backpropagation-based training is too memory demanding for their embedded assets. Using Extreme Learning Machines based on Convolutional Neural Networks might be feasible and very convenient, especially for Feature Extraction tasks. However, it requires searching for a randomly initialized topology that achieves results as good as those achieved by the backpropagated model. This work proposes a novel approach for automatically composing an Extreme Convolutional Feature Extractor, based on Neural Architecture Search and Bayesian Optimization. It was applied to the CIFAR-10 and MNIST datasets for evaluation. Two search spaces have been defined, as well as a search strategy that has been tested with two surrogate models, Gaussian Process and Random Forest. A performance estimation strategy was defined, keeping the feature set computed by the MLCommons-Tiny benchmark ResNet as a reference model. In as few as 1200 search iterations, the proposed strategy was able to achieve a topology whose extracted features scored a mean square error equal to 0.64 compared to the reference set. Further improvements are required, with a target of at least one order of magnitude decrease in mean square error for improved classification accuracy. The code is made available via GitHub to allow for the reproducibility of the results reported in this paper. [ABSTRACT FROM AUTHOR]

Identifying menopause-related breast cancer biomarkers is crucial for enhancing diagnosis, prognosis, and personalized treatment at that stage of the patient's life. In this paper, we present a comprehensive framework for extracting multiomics biomarkers specifically related to breast cancer incidence before and after menopause. Our approach integrates DNA methylation, gene expression, and copy number alteration data using a systematic pipeline encompassing data preprocessing and handling class imbalance, dimensionality reduction, and classification. The framework starts with MutSigCV for data preprocessing and ensuring data quality. The Synthetic Minority Over-sampling Technique (SMOTE) up-sampling technique is applied to address the class imbalance representation. Then, Principal Component Analysis (PCA) transforms the DNA methylation, gene expression, and copy number alteration data into a latent space. The purpose is to discard irrelevant variations and extract relevant information. Finally, a classification model is built based on the transformed multiomics data into a unified representation. The framework contributes to understanding the complex interplay between menopause and breast cancer, thereby revealing more precise diagnostic and therapeutic strategies in the future. The explainable artificial intelligence model Shapley based on the XGBoost regressor showed the power of the selected gene expressions for predicting the menopause status, and the potential biomarkers included RUNX1, PTEN, MAP3K1, and CDH1. The literature confirmed the findings. [ABSTRACT FROM AUTHOR]

Today, specific convolution neural network (CNN) models assigned to specific tasks are often used. In this article, the authors explored three models: MobileNet, EfficientNetB0, and InceptionV3 combined. The authors were interested in investigating how quickly an artificial intelligence model can be taught with limited computer resources. Three types of training bases were investigated, starting with a simple base verifying five colours, then recognizing two different orthogonal elements, followed by more complex images from different families. This research aimed to demonstrate the capabilities of the models based on training base parameters such as the number of images and epoch types. Architectures proposed by the authors in these cases were chosen based on simulation studies conducted on a virtual machine with limited hardware parameters. The proposals present the advantages and disadvantages of the different models based on the TensorFlow and Keras libraries in the Jupiter environment based on the Python programming language. An artificial intelligence model with a combination of MobileNet, proposed by Siemens, and Efficient and Inception, selected by the authors, allows for further work to be conducted on image classification, but with limited computer resources for industrial implementation on a programmable logical controller (PLC). The study showed a 90% success rate, with a learning time of 180 s. [ABSTRACT FROM AUTHOR]

Respiratory Inductance Plethysmography (RIP) is a non-invasive method for the measurement of respiratory rates and lung volumes. Accurate detection of respiratory rates and volumes is crucial for the diagnosis and monitoring of prognosis of lung diseases, for which spirometry is classically used in clinical applications. RIP has been studied as an alternative to spirometry and shown promising results. Moreover, RIP data can be analyzed through machine learning (ML)-based approaches for some other purposes, i.e., detection of apneas, work of breathing (WoB) measurement, and recognition of human activity based on breathing patterns. The goal of this study is to provide an in-depth systematic review of the scope of usage of RIP and current RIP device developments, as well as to evaluate the performance, usability, and reliability of ML-based data analysis techniques within its designated scope while adhering to the PRISMA guidelines. This work also identifies research gaps in the field and highlights the potential scope for future work. The IEEE Explore, Springer, PLoS One, Science Direct, and Google Scholar databases were examined, and 40 publications were included in this work through a structured screening and quality assessment procedure. Studies with conclusive experimentation on RIP published between 2012 and 2023 were included, while unvalidated studies were excluded. The findings indicate that RIP is an effective method to a certain extent for testing and monitoring respiratory functions, though its accuracy is lacking in some settings. However, RIP possesses some advantages over spirometry due to its non-invasive nature and functionality for both stationary and ambulatory uses. RIP also demonstrates its capabilities in ML-based applications, such as detection of breathing asynchrony, classification of apnea, identification of sleep stage, and human activity recognition (HAR). It is our conclusion that, though RIP is not yet ready to replace spirometry and other established methods, it can provide crucial insights into subjects' condition associated to respiratory illnesses. The implementation of artificial intelligence (AI) could play a potential role in improving the overall effectiveness of RIP, as suggested in some of the selected studies. [ABSTRACT FROM AUTHOR]

Genomic selection (GS) is a groundbreaking statistical machine learning method for advancing plant and animal breeding. Nonetheless, its practical implementation remains challenging due to numerous factors affecting its predictive performance. This research explores the potential of data augmentation to enhance prediction accuracy across entire datasets and specifically within the top 20% of the testing set. Our findings indicate that, overall, the data augmentation method (method A), when compared to the conventional model (method C) and assessed using Mean Arctangent Absolute Prediction Error (MAAPE) and normalized root mean square error (NRMSE), did not improve the prediction accuracy for the unobserved cultivars. However, significant improvements in prediction accuracy (evidenced by reduced prediction error) were observed when data augmentation was applied exclusively to the top 20% of the testing set. Specifically, reductions in MAAPE_20 and NRMSE_20 by 52.86% and 41.05%, respectively, were noted across various datasets. Further investigation is needed to refine data augmentation techniques for effective use in genomic prediction. [ABSTRACT FROM AUTHOR]

This study introduces a novel algorithm tailored for the precise detection of lower outliers (i.e., data points at the lower tail) in univariate datasets, which is particularly suited for scenarios with a single cluster and similar data distribution. The approach leverages a combination of transformative techniques and advanced filtration methods to efficiently segregate anomalies from normal values. Notably, the algorithm emphasizes high-precision outlier detection, ensuring minimal false positives, and requires only a few parameters for configuration. Its unsupervised nature enables robust outlier filtering without the need for extensive manual intervention. To validate its efficacy, the algorithm is rigorously tested using real-world data obtained from photovoltaic (PV) module strings with similar DC capacities, containing various outliers. The results demonstrate the algorithm's capability to accurately identify lower outliers while maintaining computational efficiency and reliability in practical applications. [ABSTRACT FROM AUTHOR]

The adequacy and efficacy of simple and hybrid machine learning and Computational Intelligence algorithms were evaluated for the classification of potential prostate cancer patients in two distinct categories, the high- and the low-risk group for PCa. The evaluation is based on randomly generated surrogate data for the biomarker PSA, considering that reported epidemiological data indicated that PSA values follow a lognormal distribution. In addition, four more biomarkers were considered, namely, PSAD (PSA density), PSAV (PSA velocity), PSA ratio, and Digital Rectal Exam evaluation (DRE), as well as patient age. Seven simple classification algorithms, namely, Decision Trees, Random Forests, Support Vector Machines, K-Nearest Neighbors, Logistic Regression, Naïve Bayes, and Artificial Neural Networks, were evaluated in terms of classification accuracy. In addition, three hybrid algorithms were developed and introduced in the present work, where Genetic Algorithms were utilized as a metaheuristic searching technique in order to optimize the training set, in terms of minimizing its size, to give optimal classification accuracy for the simple algorithms including K-Nearest Neighbors, a K-means clustering algorithm, and a genetic clustering algorithm. Results indicated that prostate cancer cases can be classified with high accuracy, even by the use of small training sets, with sizes that could be even smaller than 30% of the dataset. Numerous computer experiments indicated that the proposed training set minimization does not cause overfitting of the hybrid algorithms. Finally, an easy-to-use Graphical User Interface (GUI) was implemented, incorporating all the evaluated algorithms and the decision-making procedure. [ABSTRACT FROM AUTHOR]

Price prediction tools play a significant role in small investors' behavior. As such, this study aims to propose a method to more effectively predict stock prices in North America. Chiefly, the study addresses crucial questions related to the relevance of news and tweets in stock-price prediction and highlights the potential value of considering such parameters in algorithmic trading strategies—particularly during times of market panic. To this end, we develop innovative multi-layer perceptron (MLP) and long short-term memory (LSTM) neural networks to investigate the influence of Twitter count (TC), and news count (NC) variables on stock-price prediction under both normal and market-panic conditions. To capture the impact of these variables, we integrate technical variables with TC and NC and evaluate the prediction accuracy across different model types. We use Bloomberg Twitter count and news publication count variables in North American stock-price prediction and integrate them into MLP and LSTM neural networks to evaluate their impact during the market pandemic. The results showcase improved prediction accuracy, promising significant benefits for traders and investors. This strategic integration reflects a nuanced understanding of the market sentiment derived from public opinion on platforms like Twitter. [ABSTRACT FROM AUTHOR]

Customer churn is a significant concern, and the telecommunications industry has the largest annual churn rate of any major industry at over 30%. This study examines the use of ensemble learning models to analyze and forecast customer churn in the telecommunications business. Accurate churn forecasting is essential for successful client retention initiatives to combat regular customer churn. We used innovative and improved machine learning methods, including Decision Trees, Boosted Trees, and Random Forests, to enhance model interpretability and prediction accuracy. The models were trained and evaluated systematically by using a large dataset. The Random Forest model performed best, with 91.66% predictive accuracy, 82.2% precision, and 81.8% recall. Our results highlight how well the model can identify possible churners with the help of explainable AI (XAI) techniques, allowing for focused and timely intervention strategies. To improve the transparency of the decisions made by the classifier, this study also employs explainable artificial intelligence methods such as LIME and SHAP to illustrate the results of the customer churn prediction model. Our results demonstrate how crucial it is for customer relationship managers to implement strong analytical tools to reduce attrition and promote long-term economic viability in fiercely competitive marketplaces. This study indicates that ensemble learning models have strategic implications for improving consumer loyalty and organizational profitability in addition to confirming their performance. [ABSTRACT FROM AUTHOR]

Quantum computing has the potential to solve problems that are currently intractable to classical computers with algorithms like Quantum Phase Estimation (QPE); however, noise significantly hinders the performance of today's quantum computers. Machine learning has the potential to improve the performance of QPE algorithms, especially in the presence of noise. In this work, QPE circuits were simulated with varying levels of depolarizing noise to generate datasets of QPE output. In each case, the phase being estimated was generated with a phase gate, and each circuit modeled was defined by a randomly selected phase. The model accuracy, prediction speed, overfitting level and variation in accuracy with noise level was determined for 5 machine learning algorithms. These attributes were compared to the traditional method of post-processing and a 6x–36 improvement in model performance was noted, depending on the dataset. No algorithm was a clear winner when considering these 4 criteria, as the lowest-error model (neural network) was also the slowest predictor; the algorithm with the lowest overfitting and fastest prediction time (linear regression) had the highest error level and a high degree of variation of error with noise. The XGBoost ensemble algorithm was judged to be the best tradeoff between these criteria due to its error level, prediction time and low variation of error with noise. For the first time, a machine learning model was validated using a 2-qubit datapoint obtained from an IBMQ quantum computer. The best 2-qubit model predicted within 2% of the actual phase, while the traditional method possessed a 25% error. [ABSTRACT FROM AUTHOR]

Alzheimer's disease is a common type of neurodegenerative condition characterized by progressive neural deterioration. The anatomical changes associated with individuals affected by Alzheimer's disease include the loss of tissue in various areas of the brain. Magnetic Resonance Imaging (MRI) is commonly used as a noninvasive tool to assess the neural structure of the brain for diagnosing Alzheimer's disease. In this study, an integrated Improved Fuzzy C-means method with improved watershed segmentation was employed to segment the brain tissue components affected by this disease. These segmented features were fed into a hybrid technique for classification. Specifically, a hybrid Convolutional Neural Network–Long Short-Term Memory classifier with 14 layers was developed in this study. The evaluation results revealed that the proposed method achieved an accuracy of 98.13% in classifying segmented brain images according to different disease severities. [ABSTRACT FROM AUTHOR]

This article explores the intertwining connections among artificial intelligence, machine learning, digital transformation, and computational sustainability, detailing how these elements jointly empower citizens within a smart city framework. As technological advancement accelerates, smart cities harness these innovations to improve residents' quality of life. Artificial intelligence and machine learning act as data analysis powerhouses, making urban living more personalized, efficient, and automated, and are pivotal in managing complex urban infrastructures, anticipating societal requirements, and averting potential crises. Digital transformation transforms city operations by weaving digital technology into every facet of urban life, enhancing value delivery to citizens. Computational sustainability, a fundamental goal for smart cities, harnesses artificial intelligence, machine learning, and digital resources to forge more environmentally responsible cities, minimize ecological impact, and nurture sustainable development. The synergy of these technologies empowers residents to make well-informed choices, actively engage in their communities, and adopt sustainable lifestyles. This discussion illuminates the mechanisms and implications of these interconnections for future urban existence, ultimately focusing on empowering citizens in smart cities. [ABSTRACT FROM AUTHOR]

Cyclic air braking is a key element for ensuring safe train operation when running on a long and steep downhill railway section. In reality, the cyclic braking performance of a train is affected by its operating environment, speed and air-refilling time. Existing optimization algorithms have the problem of low learning efficiency. To solve this problem, an intelligent control method based on the deep Q-network (DQN) algorithm for heavy-haul trains running on long and steep downhill railway sections is proposed. Firstly, the environment of heavy-haul train operation is designed by considering the line characteristics, speed limits and constraints of the train pipe's air-refilling time. Secondly, the control process of heavy-haul trains running on long and steep downhill sections is described as the reinforcement learning (RL) of a Markov decision process. By designing the critical elements of RL, a cyclic braking strategy for heavy-haul trains is established based on the reinforcement learning algorithm. Thirdly, the deep neural network and Q-learning are combined to design a neural network for approximating the action value function so that the algorithm can achieve the optimal action value function faster. Finally, simulation experiments are conducted on the actual track data pertaining to the Shuozhou–Huanghua line in China to compare the performance of the Q-learning algorithm against the DQN algorithm. Our findings revealed that the DQN-based intelligent control strategy decreased the air braking distance by 2.1% and enhanced the overall average speed by more than 7%. These experiments unequivocally demonstrate the efficacy and superiority of the DQN-based intelligent control strategy. [ABSTRACT FROM AUTHOR]

Image understanding plays a pivotal role in various computer vision tasks, such as extraction of essential features from images, object detection, and segmentation. At a higher level of granularity, both semantic and instance segmentation are necessary for fully grasping a scene. In recent times, the concept of panoptic segmentation has emerged as a field of study that unifies semantic and instance segmentation. This article sheds light on the pivotal role of panoptic segmentation as a visualization tool for understanding scene components, including object detection, categorization, and precise localization of scene elements. Advancements in achieving panoptic segmentation and suggested improvements to the predicted outputs through a top-down approach are discussed. Furthermore, datasets relevant to both scene recognition and panoptic segmentation are explored to facilitate a comparative analysis. Finally, the article outlines certain promising directions in image recognition and analysis by underlining the ongoing evolution in image understanding methodologies. [ABSTRACT FROM AUTHOR]

Despite medical advancements in recent years, cardiovascular diseases (CVDs) remain a major factor in rising mortality rates, challenging predictions despite extensive expertise. The healthcare sector is poised to benefit significantly from harnessing massive data and the insights we can derive from it, underscoring the importance of integrating machine learning (ML) to improve CVD prevention strategies. In this study, we addressed the major issue of class imbalance in the Behavioral Risk Factor Surveillance System (BRFSS) 2021 heart disease dataset, including personal lifestyle factors, by exploring several resampling techniques, such as the Synthetic Minority Oversampling Technique (SMOTE), Adaptive Synthetic Sampling (ADASYN), SMOTE-Tomek, and SMOTE-Edited Nearest Neighbor (SMOTE-ENN). Subsequently, we trained, tested, and evaluated multiple classifiers, including logistic regression (LR), decision trees (DTs), random forest (RF), gradient boosting (GB), XGBoost (XGB), CatBoost, and artificial neural networks (ANNs), comparing their performance with a primary focus on maximizing sensitivity for CVD risk prediction. Based on our findings, the hybrid resampling techniques outperformed the alternative sampling techniques, and our proposed implementation includes SMOTE-ENN coupled with CatBoost optimized through Optuna, achieving a remarkable 88% rate for recall and 82% for the area under the receiver operating characteristic (ROC) curve (AUC) metric. [ABSTRACT FROM AUTHOR]

Childhood stroke can lead to lifelong disability. Developing algorithms for timely recognition of clinical and paraclinical signs is crucial to ensure prompt stroke diagnosis and minimize decision-making time. This study aimed to characterize clinical and paraclinical symptoms of childhood and neonatal stroke as relevant diagnostic criteria encountered in clinical practice, in order to develop algorithms for prompt stroke diagnosis. The analysis included data from 402 pediatric case histories from 2010 to 2016 and 108 prospective stroke cases from 2017 to 2020. Stroke cases were predominantly diagnosed in newborns, with 362 (71%, 95% CI 68.99–73.01) cases occurring within the first 28 days of birth, and 148 (29%, 95% CI 26.99–31.01) cases occurring after 28 days. The findings of the study enable the development of algorithms for timely stroke recognition, facilitating the selection of optimal treatment options for newborns and children of various age groups. Logistic regression serves as the basis for deriving these algorithms, aiming to initiate early treatment and reduce lifelong morbidity and mortality in children. The study outcomes include the formulation of algorithms for timely recognition of newborn stroke, with plans to adopt these algorithms and train a fuzzy classifier-based diagnostic model using machine learning techniques for efficient stroke recognition. [ABSTRACT FROM AUTHOR]

In this study, we investigate Quantum Long Short-Term Memory and Quantum Gated Recurrent Unit integrated with Variational Quantum Circuits in modeling complex dynamical systems, including the Van der Pol oscillator, coupled oscillators, and the Lorenz system. We implement these advanced quantum machine learning techniques and compare their performance with traditional Long Short-Term Memory and Gated Recurrent Unit models. The results of our study reveal that the quantum-based models deliver superior precision and more stable loss metrics throughout 100 epochs for both the Van der Pol oscillator and coupled harmonic oscillators, and 20 epochs for the Lorenz system. The Quantum Gated Recurrent Unit outperforms competing models, showcasing notable performance metrics. For the Van der Pol oscillator, it reports MAE 0.0902 and RMSE 0.1031 for variable x and MAE 0.1500 and RMSE 0.1943 for y; for coupled oscillators, Oscillator 1 shows MAE 0.2411 and RMSE 0.2701 and Oscillator 2 MAE is 0.0482 and RMSE 0.0602; and for the Lorenz system, the results are MAE 0.4864 and RMSE 0.4971 for x, MAE 0.4723 and RMSE 0.4846 for y, and MAE 0.4555 and RMSE 0.4745 for z. These outcomes mark a significant advancement in the field of quantum machine learning. [ABSTRACT FROM AUTHOR]

Background. Brain–machine interfaces (BMIs) offer users the ability to directly communicate with digital devices through neural signals decoded with machine learning (ML)-based algorithms. Spiking Neural Networks (SNNs) are a type of Artificial Neural Network (ANN) that operate on neural spikes instead of continuous scalar outputs. Compared to traditional ANNs, SNNs perform fewer computations, use less memory, and mimic biological neurons better. However, SNNs only retain information for short durations, limiting their ability to capture long-term dependencies in time-variant data. Here, we propose a novel spike-weighted SNN with spiking long short-term memory (swSNN-SLSTM) for a regression problem. Spike-weighting captures neuronal firing rate instead of membrane potential, and the SLSTM layer captures long-term dependencies. Methods. We compared the performance of various ML algorithms during decoding directional movements, using a dataset of microelectrode recordings from a macaque during a directional joystick task, and also an open-source dataset. We thus quantified how swSNN-SLSTM performed compared to existing ML models: an unscented Kalman filter, LSTM-based ANN, and membrane-based SNN techniques. Result. The proposed swSNN-SLSTM outperforms both the unscented Kalman filter, the LSTM-based ANN, and the membrane based SNN technique. This shows that incorporating SLSTM can better capture long-term dependencies within neural data. Also, our proposed swSNN-SLSTM algorithm shows promise in reducing power consumption and lowering heat dissipation in implanted BMIs. [ABSTRACT FROM AUTHOR]

Gout is one of the most painful diseases in the world. Accurate classification of gout is crucial for diagnosis and treatment which can potentially save lives. However, the current methods for classifying gout periods have demonstrated poor performance and have received little attention. This is due to a significant data imbalance problem that affects the learning attention for the majority and minority classes. To overcome this problem, a resampling method called ENaNSMOTE-Tomek link is proposed. It uses extended natural neighbors to generate samples that fall within the minority class and then applies the Tomek link technique to eliminate instances that contribute to noise. The model combines the ensemble 'bagging' technique with the proposed resampling technique to improve the quality of generated samples. The performance of individual classifiers and hybrid models on an imbalanced gout dataset taken from the electronic medical records of a hospital is evaluated. The results of the classification demonstrate that the proposed strategy is more accurate than some imbalanced gout diagnosis techniques, with an accuracy of 80.87% and an AUC of 87.10%. This indicates that the proposed algorithm can alleviate the problems caused by imbalanced gout data and help experts better diagnose their patients. [ABSTRACT FROM AUTHOR]

Anomaly detection in data streams (and particularly time series) is today a vitally important task. Machine learning algorithms are a common design for achieving this goal. In particular, deep learning has, in the last decade, proven to be substantially more accurate than shallow learning in a wide variety of machine learning problems, and deep anomaly detection is very effective for point anomalies. However, deep semi-supervised contextual anomaly detection (in which anomalies within a time series are rare and none at all occur in the algorithm's training data) is a more difficult problem. Hybrid anomaly detectors (a "normal model" followed by a comparator) are one approach to these problems, but the separate loss functions for the two components can lead to inferior performance. We investigate a novel synthetic-example oversampling technique to harmonize the two components of a hybrid system, thus improving the anomaly detector's performance. We evaluate our algorithm on two distinct problems: identifying pipeline leaks and patient-ventilator asynchrony. [ABSTRACT FROM AUTHOR]

Endometriosis (EM) is a chronic inflammatory estrogen-dependent disorder that affects 10% of women worldwide. It affects the female reproductive tract and its resident microbiota, as well as distal body sites that can serve as surrogate markers of EM. Currently, no single definitive biomarker can diagnose EM. For this pilot study, we analyzed a cohort of 21 patients with endometriosis and infertility-associated conditions. A microbiome dataset was created using five sample types taken from the reproductive and gastrointestinal tracts of each patient. We evaluated several machine learning algorithms for EM detection using these features. The characteristics of the dataset were derived from endometrial biopsy, endometrial fluid, vaginal, oral, and fecal samples. Despite limited data, the algorithms demonstrated high performance with respect to the F1 score. In addition, they suggested that disease diagnosis could potentially be improved by using less medically invasive procedures. Overall, the results indicate that machine learning algorithms can be useful tools for diagnosing endometriosis in low-resource settings where data availability and availability are limited. We recommend that future studies explore the complexities of the EM disorder using artificial intelligence and prediction modeling to further define the characteristics of the endometriosis phenotype. [ABSTRACT FROM AUTHOR]

The Recurrent Neural Networks (RNNs) are an essential class of supervised learning algorithms. Complex tasks like speech recognition, machine translation, sentiment classification, weather prediction, etc., are now performed by well-trained RNNs. Local or cloud-based GPU machines are used to train them. However, inference is now shifting to miniature, mobile, IoT devices and even micro-controllers. Due to their colossal memory and computing requirements, mapping RNNs directly onto resource-constrained platforms is arcane and challenging. The efficacy of edge-intelligent RNNs (EI-RNNs) must satisfy both performance and memory-fitting requirements at the same time without compromising one for the other. This study's aim was to provide an empirical evaluation and optimization of historic as well as recent RNN architectures for high-performance and low-memory footprint goals. We focused on Human Activity Recognition (HAR) tasks based on wearable sensor data for embedded healthcare applications. We evaluated and optimized six different recurrent units, namely Vanilla RNNs, Long Short-Term Memory (LSTM) units, Gated Recurrent Units (GRUs), Fast Gated Recurrent Neural Networks (FGRNNs), Fast Recurrent Neural Networks (FRNNs), and Unitary Gated Recurrent Neural Networks (UGRNNs) on eight publicly available time-series HAR datasets. We used the hold-out and cross-validation protocols for training the RNNs. We used low-rank parameterization, iterative hard thresholding, and spare retraining compression for RNNs. We found that efficient training (i.e., dataset handling and preprocessing procedures, hyperparameter tuning, and so on, and suitable compression methods (like low-rank parameterization and iterative pruning) are critical in optimizing RNNs for performance and memory efficiency. We implemented the inference of the optimized models on Raspberry Pi. [ABSTRACT FROM AUTHOR]

The distributed denial of service (DDoS) attack is one of the most pernicious threats in cyberspace. Catastrophic failures over the past two decades have resulted in catastrophic and costly disruption of services across all sectors and critical infrastructure. Machine-learning-based approaches have shown promise in developing intrusion detection systems (IDSs) for detecting cyber-attacks, such as DDoS. Herein, we present a solution to detect DDoS attacks through an ensemble-based machine learning approach that combines supervised and unsupervised machine learning ensemble frameworks. This combination produces higher performance in detecting known DDoS attacks using supervised ensemble and for zero-day DDoS attacks using an unsupervised ensemble. The unsupervised ensemble, which employs novelty and outlier detection, is effective in identifying prior unseen attacks. The ensemble framework is tested using three well-known benchmark datasets, NSL-KDD, UNSW-NB15, and CICIDS2017. The results show that ensemble classifiers significantly outperform single-classifier-based approaches. Our model with combined supervised and unsupervised ensemble models correctly detects up to 99.1% of the DDoS attacks, with a negligible rate of false alarms. [ABSTRACT FROM AUTHOR]

High-value timber species with economic and ecological importance are usually distributed at very low densities, such that accurate knowledge of the location of these trees within a forest is critical for forest management practices. Recent technological developments integrating unmanned aerial vehicle (UAV) imagery and deep learning provide an efficient method for mapping forest attributes. In this study, we explored the applicability of high-resolution UAV imagery and a deep learning algorithm to predict the distribution of high-value deciduous broadleaf tree crowns of Japanese oak (Quercus crispula) in an uneven-aged mixed forest in Hokkaido, northern Japan. UAV images were collected in September and October 2022 before and after the color change of the leaves of Japanese oak to identify the optimal timing of UAV image collection. RGB information extracted from the UAV images was analyzed using a ResU-Net model (U-Net model with a Residual Network 101 (ResNet101), pre-trained on large ImageNet datasets, as backbone). Our results, confirmed using validation data, showed that reliable F1 scores (>0.80) could be obtained with both UAV datasets. According to the overlay analyses of the segmentation results and all the annotated ground truth data, the best performance was that of the model with the October UAV dataset (F1 score of 0.95). Our case study highlights a potential methodology to offer a transferable approach to the management of high-value timber species in other regions. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have proven to be a very efficient class of machine learning (ML) architectures for handling multidimensional data by maintaining data locality, especially in the field of computer vision. Data pooling, a major component of CNNs, plays a crucial role in extracting important features of the input data and downsampling its dimensionality. Multidimensional pooling, however, is not efficiently implemented in existing ML algorithms. In particular, quantum machine learning (QML) algorithms have a tendency to ignore data locality for higher dimensions by representing/flattening multidimensional data as simple one-dimensional data. In this work, we propose using the quantum Haar transform (QHT) and quantum partial measurement for performing generalized pooling operations on multidimensional data. We present the corresponding decoherence-optimized quantum circuits for the proposed techniques along with their theoretical circuit depth analysis. Our experimental work was conducted using multidimensional data, ranging from 1-D audio data to 2-D image data to 3-D hyperspectral data, to demonstrate the scalability of the proposed methods. In our experiments, we utilized both noisy and noise-free quantum simulations on a state-of-the-art quantum simulator from IBM Quantum. We also show the efficiency of our proposed techniques for multidimensional data by reporting the fidelity of results. [ABSTRACT FROM AUTHOR]

Background: The development and application of machine learning (ML) methods have become so fast that almost nobody can follow their developments in every detail. It is no wonder that numerous errors and inconsistencies in their usage have also spread with a similar speed independently from the tasks: regression and classification. This work summarizes frequent errors committed by certain authors with the aim of helping scientists to avoid them. Methods: The principle of parsimony governs the train of thought. Fair method comparison can be completed with multicriteria decision-making techniques, preferably by the sum of ranking differences (SRD). Its coupling with analysis of variance (ANOVA) decomposes the effects of several factors. Earlier findings are summarized in a review-like manner: the abuse of the correlation coefficient and proper practices for model discrimination are also outlined. Results: Using an illustrative example, the correct practice and the methodology are summarized as guidelines for model discrimination, and for minimizing the prediction errors. The following factors are all prerequisites for successful modeling: proper data preprocessing, statistical tests, suitable performance parameters, appropriate degrees of freedom, fair comparison of models, and outlier detection, just to name a few. A checklist is provided in a tutorial manner on how to present ML modeling properly. The advocated practices are reviewed shortly in the discussion. Conclusions: Many of the errors can easily be filtered out with careful reviewing. Every authors' responsibility is to adhere to the rules of modeling and validation. A representative sampling of recent literature outlines correct practices and emphasizes that no error-free publication exists. [ABSTRACT FROM AUTHOR]

We present a novel technique for pruning called activation-based pruning to effectively prune fully connected feedforward neural networks for multi-object classification. Our technique is based on the number of times each neuron is activated during model training. We compare the performance of activation-based pruning with a popular pruning method: magnitude-based pruning. Further analysis demonstrated that activation-based pruning can be considered a dimensionality reduction technique, as it leads to a sparse low-rank matrix approximation for each hidden layer of the neural network. We also demonstrate that the rank-reduced neural network generated using activation-based pruning has better accuracy than a rank-reduced network using principal component analysis. We provide empirical results to show that, after each successive pruning, the amount of reduction in the magnitude of singular values of each matrix representing the hidden layers of the network is equivalent to introducing the sum of singular values of the hidden layers as a regularization parameter to the objective function. [ABSTRACT FROM AUTHOR]