Search

The application of blockchain technology is growing rapidly, which has aroused great attention in the academic and industrial fields. Based on blockchain 2.0, Ethereum is a mainstream smart contract development and operation platform. The trading process of Ethereum users is facing a serious threat of financial fraud. In particular, the Ponzi scheme is a classic form of fraud. Relevant works have investigated the issue of Ponzi schemes smart contract detection on Ethereum based on machine learning approaches. Nevertheless, the detection approaches still fall short in dealing with the big data-space Ponzi scheme smart contract detection application based on the class-imbalanced training data. We propose PSD-OL, a Ponzi schemes detection approach based on oversampling-based Long Short-Term Memory (LSTM) for smart contracts in this paper. PSD-OL takes the contract account features and the contract code features together into consideration. Oversampling technique is utilized to fill the class-imbalanced Ponzi scheme smart contracts' sample feature data. An LSTM model is trained by learning from the feature data for future Ponzi scheme detection. Experimental results conducted on the well-known XBlock dataset demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Liver diseases represent a significant healthcare challenge, impacting millions globally and posing complexities in diagnosis. To address this global health concern, this paper introduces a groundbreaking enhancement to the Kepler Optimization Algorithm, termed I-KOA, designed specifically for feature selection in high-dimensional datasets. By harnessing the synergies of Opposition-Based Learning and a Local Escaping Operator grounded in the k-nearest Neighbor (kNN) classifier, I-KOA asserts itself as a potent tool for local exploitation, balanced exploration, and evasion of local optima. To our knowledge, this is the first work to exploit KOA as a feature selection method. Pioneering the utilization of KOA as a feature selection method, the paper rigorously tests I-KOA in two extensive experiments, tackling the complex CEC'22 benchmark suite functions and the intricate landscape of five liver disease datasets. Results underscore I-KOA's unparalleled performance, validated through the Friedman test, where it surpasses seven rival optimization algorithms. Achieving an outstanding overall classification accuracy of 93.46%, Feature selection size of 0.1042, sensitivity of 97.46%, precision of 94.37%, and F1-score of 90.35% across the liver disease datasets, I-KOA's randomized algorithm ensures robust feature selection, striking a compelling balance between subset size and classification efficacy. Acknowledging computational demands and generalization nuances, I-KOA is a formidable tool ready to revolutionize medical diagnosis and decision support systems. The open source codes of the proposed I-KOA are available at https://www.mathworks.com/matlabcentral/fileexchange/161376-improved-kepler-optimization-algorithm. • This paper encompasses a proficient I-KOA algorithm based on OBL and LEO methods. • A new optimized feature selection model for liver disease classification using five datasets. • We employ comprehensive analysis metrics to thoroughly assess the I-KOA algorithm's efficacy. • I-KOA stands out by surpassing its competitors, attesting to its remarkable performance. [ABSTRACT FROM AUTHOR]

An Electrical Submersible Pump (ESP) is an important equipment used in the industry for lifting liquids in various types of wells. An ESP is widely used in the oil industry for offshore exploration. Detecting a faulty ESP before installation is a predictive maintenance measure in order to extend its operational time. Machine learning fault diagnosis is an effective way for performing this task. Machine learning fault diagnosis algorithms are highly dependent of the availability of an appropriate problem dataset. This paper describes in detail the problem of ESP fault diagnosis and the ESPset dataset, a real-world and public dataset for vibration-based fault diagnosis of electrical submersible pumps used on offshore oil exploration. In addition, the paper also proposes an experimental framework for adequately comparing research works based on the ESPset dataset and defines benchmark classifiers and respective results as referential to the fault diagnosis research community. The framework considers the fact that some subset of samples are not drawn independently, and therefore, proposes a cross-validation sampling strategy that mitigates the similarity bias among samples. Indeed, this work shows that a conventional k-fold cross-validation may lead to a clear overestimation of the average performance. This fact is supported by results which show that the best classification model drops from a mean F-measure of 0.887 to 0.733 when removing the similarity bias from the data. • The Electrical Submersible Pumps (ESP) fault diagnosis problem description: a broader and detailed specification. • The ESPset vibration dataset is published: 6032 instances of ESP vibration tests. • An experimental framework to avoid dependence between the training and test phases of the learning task. • The ESPset benchmark classifiers are presented. • The results of the experimental study using the proposed framework and the benchmark classifiers are shown. [ABSTRACT FROM AUTHOR]

Feature selection (FS) is a crucial step in machine learning and data mining projects. It aims to remove redundant and uncorrelated features, thus improving the accuracy of models. However, it can be challenging to select the optimal features, especially for datasets with many features. This paper proposes the Binary Meerkat Optimization Algorithm (BMOA) to address the issue of FS. The BMOA selects the most related and useful features. Additionally, an improved BMOA (IBMOA) is introduced, which enhances exploration and exploitation capabilities by incorporating the Periodic Mode Boundary Handling (PMBH) strategy and the Local Search (LS) process. This helps to reduce dimensionality and enhance classification accuracy. To evaluate the significance of selected features, two widely used classifiers, namely k -nearest Neighbor (k -NN) and Support Vector Machine (SVM), were used as quality raters. The proposed IBMOA and the original BMOA were compared on 21 multi-scale and multi-faceted benchmark datasets. The binary structures of eight contemporary algorithms, including Binary Artificial Bee Colony (BABC), Binary Grey Wolf Optimization (BGWO), Binary Sailfish Optimizer (BSFO), Binary Particle Swarm Optimizer (BPSO), Binary Bird Swarm Algorithm (BBSA), Binary Whale Optimization Algorithm (BWOA), Binary Grasshopper Optimization Algorithm (BGOA), and Binary Bat Algorithm (BBA), were also analyzed and compared. The proposed IBMOA algorithm outperforms competitors on small and large datasets, obtaining optimal solutions within a reasonable timeframe. This is true for the proposed IBMOA, which has been statistically proven to be highly competitive using the Wilcoxon rank sum test (with alpha=0.05). [ABSTRACT FROM AUTHOR]

In an effort to reduce pesticide use, agronomists and computer scientists have joined forces to develop site-specific weed detection and classification systems. These systems aim to recognize and locate weed species within a crop field, using precision equipment to apply required herbicides timely and only where needed, with the objective of reducing the sprayable surface required to eliminate the given weed and protect the crop, with both economic and environmental benefits. Yet, with climate change on the rise, common weeds are expected to undergo some changes to adapt to their environment, possibly with new or invasive weeds spreading to areas where they did not exist before. These changes (often morphological) as well as new invasions need to be taken into account by future classifiers and detection algorithms to ensure system robustness and adaptation to new habitats/climate dynamics. This paper proposes a set of experiments evaluating the use of transfer learning and zero-shot learning for weed classification using our novel TomatoWeeds dataset. Residual networks of variable depth, pretrained on the Imagenet and/or DeepWeeds datasets were evaluated. A ResNet50 pretrained on both datasets and fine-tuned on the TomatoWeeds dataset performed best, returning a holdout set accuracy of 77.8%, showing the advantageous use of transfer learning in this domain. Zero-shot learning, using both embeddings of images and morphological and habitat text-based descriptions, is implemented to test the ability of machine learning pipelines of recognizing unseen classes at test time (which may arise e.g. due to changing climate dynamics), a learning task in which the field (and our experiments) are still far from satisfactory results. Further research could benefit from larger weed-specific datasets for transfer learning as well as deeper network architectures to improve model performance. The projection-based ZSL could also benefit from larger datasets and new zero-shot learning architectures in hope that unseen classes are accurately projected. • Use of the TomatoWeed dataset; containing UAV images of tomato fields. • Three individual weed species are labeled using single pixel annotation. • Different transfer learning configuration were tested. • Resnet50 pretrained on Imagenet and DeepWeeds was deemed best performing. • Zero-shot learning approach built to detect unseen weed classes at testing. [ABSTRACT FROM AUTHOR]

The vast majority of machine learning research focuses on improving the correctness of the outcomes (i.e., accuracy, error-rate, and other metrics). However, the negative impact of machine learning outcomes can be substantial if the consequences marginalize certain groups of data, especially if certain groups of people are the ones being discriminated against. Thus, recent papers try to tackle the unfair treatment of certain groups of data (humans), but mostly focus on only one sensitive feature with binary values. In this paper, we propose an ensemble boosting FairBoost that takes into consideration fairness as well as accuracy to mitigate unfairness in classification tasks during the model training process. This method tries to close the gap between proposed approaches and real-world applications, where there is often more than one sensitive feature that contains multiple categories. The proposed approach checks the bias and corrects it through the iteration of building the boosted ensemble. The proposed FairBoost is tested within the experimental setting and compared to similar existing algorithms. The results on different datasets and settings show no significant changes in the overall quality of classification, while the fairness of the outcomes is vastly improved. [ABSTRACT FROM AUTHOR]

This paper presents two novel data reduction algorithms that combine instance selection and feature selection methods to simultaneously reduce the number of records and features. These algorithms are rigorously defined and thoroughly explained. They are applied to reduce twelve data sets selected from a commonly available database repository and single representative of big database. Their reduction efficiency is tested using three well-known machine learning models: classification and regression tree, convolutional neural network, and support vector machines. The results are extensively analyzed, comparing the classification performance of the models on the reduced data sets with that on the original ones. The computational complexity of algorithms is theoretically estimated and practically verified. The obtained performance results are compared with those available in the literature. It is shown that implementing the proposed data reduction algorithms consistently improves model accuracy in the vast majority of classification scenarios. [ABSTRACT FROM AUTHOR]

Within the contemporary dairy industry, the effective monitoring of cheese ripeness constitutes a critical yet challenging task. This paper proposes the first public dataset encompassing images of cheese wheels that depict various products at distinct stages of ripening and introduces an innovative hybrid approach, integrating machine learning and computer vision techniques to automate the detection of cheese ripeness. By leveraging deep learning and shallow learning techniques, the proposed method endeavors to overcome the limitations associated with conventional assessment methodologies. It aims to provide automation, precision, and consistency in the evaluation of cheese ripeness, delving into a hierarchical classification for the simultaneous classification of distinct cheese types and ripeness levels and presenting a comprehensive solution to enhance the efficiency of the cheese production process. By employing a lightweight hierarchical feature aggregation methodology, this investigation navigates the intricate landscape of preprocessing steps, feature selection, and diverse classifiers. We report a noteworthy achievement, attaining a best F-measure score of 0.991 through the merging of features extracted from EfficientNet and DarkNet-53, opening the field to concretely address the complexity inherent in cheese quality assessment. [ABSTRACT FROM AUTHOR]

Artificial neural networks (ANNs), inspired by the interconnection of real neurons, have achieved unprecedented success in various fields such as computer vision and natural language processing. Recently, a novel mathematical ANN model, known as the dendritic neuron model (DNM), has been proposed to address nonlinear problems by more accurately reflecting the structure of real neurons. However, the single-output design limits its capability to handle multi-output tasks, significantly lowering its applications. In this paper, we propose a novel multi-in and multi-out dendritic neuron model (MODN) to tackle multi-output tasks. Our core idea is to introduce a filtering matrix to the soma layer to adaptively select the desired dendrites to regress each output. Because such a matrix is designed to be learnable, MODN can explore the relationship between each dendrite and output to provide a better solution to downstream tasks. We also model a telodendron layer into MODN to simulate better the real neuron behavior. Importantly, MODN is a more general and unified framework that can be naturally specialized as the DNM by customizing the filtering matrix. To explore the optimization of MODN, we investigate both heuristic and gradient-based optimizers and introduce a two-step training method for MODN. Extensive experimental results performed on 11 datasets on both binary and multi-class classification tasks demonstrate the effectiveness of MODN, with respect to accuracy, convergence, and generality. [ABSTRACT FROM AUTHOR]

Structurally diverse decision trees are important for knowledge discovery and classification/prediction accuracy. Over the years, researchers have devoted much effort to the development of algorithms to increase diversity among the trees within an ensemble. While Kappa is commonly used to measure diversity among the decision trees, it does not measure the ability of the tree building algorithms to introduce diversity. Further, Kappa does not consider the structural diversity amongst the trees. Instead, Kappa measures the diversity of the predictions made from the trees produced, and are dependent on the datasets used. This paper presents a novel data-independent metric, called R index, for measuring the diversity that can be introduced by a decision forest algorithm without building the entire decision forest. The proposed measure is applied to five well-known algorithms that involve bagging and random subspacing. An efficient practical approach for calculating the R index empirically – R finder – is also proposed, and is implemented. Both R finder and Kappa were applied to thirty-two publicly available benchmark datasets under various algorithms to estimate the resulting diversity. The results indicate a generally strong negative correlation between R finder and Kappa, implying that R finder is effective at estimating the diversity of trees without the added computational costs associated with calculating Kappa. [ABSTRACT FROM AUTHOR]

Abstract In this paper, we investigate the unsupervised domain transfer learning in which there is no label in the target samples whereas the source samples are all labeled. We use the transformation matrix to transfer both target and source samples to a common subspace where they have the same distribution and each target sample in the transformed space is constructed of a linear combination of the source samples. To preserve the local and global structure of the samples in the transferred domain, the low-rank and sparse constraints are imposed on the reconstruction coefficient matrix. In this paper, in order to consider the discriminative ability of the target and source samples, the information content of the reconstruction coefficient matrix is utilized. To capture the discriminative ability of the target samples, it is assumed that the class labels of the source samples which are linearly incorporated in constructing a target sample should be the same. Based on this assumption, it is assured that the target samples are well distributed over the transferred domain. To handle this, we utilize the linear entropy to measure the discriminant power of the target domain. This term considers the discriminative ability of the target samples without using their hidden labels. Also, to assess the discriminative ability of source samples, we use max-margin classifier where the kernel matrix is defined by using the reconstruction coefficient matrix. To evaluate the proposed approach, it is applied on MSRC, VOC 2007, CMU PIE, Office, Caltech-256, Extended Yale B and two imbalanced datasets. The experimental results show that our proposed approach outperforms its competitors. Highlights • A new approach in unsupervised domain transfer learning is proposed. • The low rank and sparse constraints are imposed on the reconstruction matrix. • The discriminative ability of the target and source samples is captured. • The information content of the reconstruction coefficient matrix is utilized. [ABSTRACT FROM AUTHOR]

Abstract This paper addresses the problem of a finite set of entities which are required to achieve a goal within a predefined deadline. For example, a group of students is supposed to submit a homework by a specified cutoff. Further, we are interested in predicting which entities will achieve the goal within the deadline. The predictive models are built based only on the data from that population. The predictions are computed at various time instants by taking into account updated data about the entities. The first contribution of the paper is a formal description of the problem. The important characteristic of the proposed method for model building is the use of the properties of entities that have already achieved the goal. We call such an approach “Self-Learning”. Since typically only a few entities have achieved the goal at the beginning and their number gradually grows, the problem is inherently imbalanced. To mitigate the curse of imbalance, we improved the Self-Learning method by tackling information loss and by several sampling techniques. The original Self-Learning and the modifications have been evaluated in a case study for predicting submission of the first assessment in distance higher education courses. The results show that the proposed improvements outperform the specified two base-line models and the original Self-Learner, and also that the best results are achieved if domain-driven techniques are utilised to tackle the imbalance problem. We also showed that these improvements are statistically significant using Wilcoxon signed rank test. [ABSTRACT FROM AUTHOR]

Abstract In this paper, a sentiment classification model is proposed to address two predominant issues in sentiment classification, namely domain-sensitive and data imbalance. Since words may embed distinct sentiment polarities in different contexts, sentiment classification is widely contended as a domain-sensitive task. Accordingly, this paper draws on label propagation to induce universal and domain-specific sentiment lexicons and builds a domain-adaptive sentiment classification model that incorporates universal and domain-specific knowledge into a unified learning framework. On the flip side, sentiment-related corpuses are usually formed with skewed polarity distribution because individuals tend to share similar assessment criteria on a given object and hence their sentiment polarities toward the same object are likely to be similar. We endeavor to address such imbalanced data problem by advancing a novel over-sampling technique. Unlike existing over-sampling approaches that generate minority-class samples from numerical feature space, the proposed sampling method directly creates synthetic texts from word spaces. Several experiments are conducted to verify the effectiveness of the proposed lexicon generation method, learning framework, and over-sampling method. Results show that the induced sentiment lexicons are interpretable and the proposed model is found to be effective for imbalanced and domain-specific text sentiment classification. [ABSTRACT FROM AUTHOR]

Community structure is one of the most important inherent properties in complex networks. Generally, nodes in each community span a different subspace in the geometric space. However, the subspace based methods are not well studied, and the existing related algorithms are sensitive to the perturbations in networks, thus lack the ability to capture the robust subspace segmentation accurately. In this paper we propose a low-rank subspace learning based network community detection algorithm. The contribution of this paper is to find the lowest rank representation of all node vectors jointly in the geometric space using low-rank decomposition strategy. Then the final robust low-rank subspace based communities are obtained from the lowest rank representations. Low-rank decomposition is better at capturing global structure of networks, thus our algorithm is more robust to perturbations in networks and has a better performance on discriminating the community boundaries. Experimental results on both synthetic benchmarks and real-world networks demonstrate that the proposed algorithm outperforms the state-of-the-art community detection approaches and the results basically preserve the actual community structure, especially on networks with vague community structures. [ABSTRACT FROM AUTHOR]

Linear discriminant analysis (LDA) is a powerful supervised dimensionality reduction method for analysing high-dimensional data. However, LDA cannot use locality information in data, which makes LDA degrade dramatically in performance on multimodal data. A number of LDA variants have been proposed to exploit locality information in data, including subclass-based LDAs. We discover a problem with these variants, which is that subclasses are selected on a within-class basis without considering other classes. This causes the loss of important information at class boundaries. In this paper, we present a novel variant of subclass-based LDA, Global Subclass Discriminant Analysis (GSDA). Unlike other subclass-based LDAs, GSDA selects subclasses from global clusters that may cross class boundaries, thus utilising within-class information and between-class information. More specifically, GSDA applies an effective clustering algorithm to the whole data to construct global clusters. It then utilises the local structure refining strategy on these global clusters to construct subclasses. Finally, GSDA learns a representative data subspace by maximising inter-subclass distance and minimising intra-subclass distance simultaneously. GSDA is extensively evaluated on a wide range of public datasets through comparison with the state-of-the-art LDA algorithms. Experimental results demonstrate its superiority in terms of accuracy and run times. • Linear Discriminant Analysis (LDA) is ineffective in processing multimodal data. • Global Subclass Discriminant Analysis (GSDA) is proposed to solve the above problem. • Local Structure Refining Strategy (LSRS) is proposed in GSDA. • LSRS allows GSDA separating classes at boundaries and obtain local information. • Extensive experiments demonstrate that GSDA outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Automatic speech recognition (ASR) has recently become an important challenge when using deep learning (DL). It requires large-scale training datasets and high computational and storage resources. Moreover, DL techniques and machine learning (ML) approaches in general, hypothesize that training and testing data come from the same domain, with the same input feature space and data distribution characteristics. This assumption, however, is not applicable in some real-world artificial intelligence (AI) applications. Moreover, there are situations where gathering real data is challenging, expensive, or rarely occurring, which cannot meet the data requirements of DL models. deep transfer learning (DTL) has been introduced to overcome these issues, which helps develop high-performing models using real datasets that are small or slightly different but related to the training data. This paper presents a comprehensive survey of DTL-based ASR frameworks to shed light on the latest developments and helps academics and professionals understand current challenges. Specifically, after presenting the DTL background, a well-designed taxonomy is adopted to inform the state-of-the-art. A critical analysis is then conducted to identify the limitations and advantages of each framework. Moving on, a comparative study is introduced to highlight the current challenges before deriving opportunities for future research. • ASR can be improved through language model, acoustic model, and multi-task learning. • Fine-tuning and domain adaptation are effective transfer learning methods in ASR. • DTL can enhance ASR using other domains, such as Cross-language and cross-corpus tasks. • DTL shows promise in improving ASR accuracy nor medical diagnosis. • DTL plays a role in enhancing ASR security against both white and black box attacks. [ABSTRACT FROM AUTHOR]

Multiobjective evolutionary algorithms (MOEAs) have gained much attention due to their high effectiveness and efficiency in solving multiobjective optimization problems (MOPs). However, when solving MOPs, it is important but difficult to maintain a good balance of exploration and exploitation. In addition, some reference point based MOEAs with fixed reference points perform poorly on MOPs with irregular frontiers. Therefore, this paper proposes a new multistrategy multiobjective differential evolutionary (DE) algorithm, named RLMMDE. In RLMMDE, a multistrategy and multicrossover DE optimizer is utilized to alleviate the exploration and exploitation dilemma. An adaptive reference point activation mechanism based on RL is proposed to activate the adaptive adjustment of reference points. Moreover, a reference point adaptation method is proposed to improve the performance of RLMMDE on irregular frontier problems. Experimental results of RLMMDE tested on some benchmark test suites (i.e., ZDT, DTLZ, UF, WFG, and LSMOP) and two practical mixed-variable optimization problems show that the algorithm outperforms some advanced MOEAs. [ABSTRACT FROM AUTHOR]

Existing research literally evaluates model fairness over limited observed data. In practice, however, factors such as maliciously crafted examples and naturally corrupted examples often appear in real-world data collection. This severely limits the reliability of bias removal methods, inhibits fairness improvement in long-term learning systems, and probes to study accuracy-related robustness. Therefore, we ask: How adversarial examples will skew model fairness? In this paper, we investigate the vulnerability of individual fairness and group fairness to adversarial attacks. We further propose a general adversarial fairness attack framework capable of twisting model bias through a small subset of adversarial examples. We formulate this problem as an optimization problem: maximizing the model bias with the constraint of the number of adversarial examples and the perturbation scale. Our approach finds the most vulnerable examples to model fairness based on the estimated distance from examples to the decision boundary and demographic information. The experimental results 1 1 https://github.com/TaocsZhang/Fairness-Attack-via-Adversarial-Examples. show that model fairness is easily skewed by a small number of adversarial examples. • Vulnerability of Model Fairness: We are the first to study the vulnerability of model fairness from the view of adversarial examples. • Adversarial Fairness Attacks: We formulate the adversarial fairness attack as a general optimization problem. • Empirical Benefits: We empirically evaluate proposed methods across four real datasets. [ABSTRACT FROM AUTHOR]

It is difficult for e-learners to make decisions on how to learn when they are facing with a large amount of learning resources, especially when they have to balance available limited learning time and multiple learning objectives in various learning scenarios. This research presented in this paper addresses this challenge by proposing a new multi-constraint learning path recommendation algorithm based on knowledge map. The main contributions of the paper are as follows. Firstly, two hypotheses on e-learners’ different learning path preferences for four different learning scenarios (initial learning, usual review, pre-exam learning and pre-exam review) are verified through questionnaire-based statistical analysis. Secondly, according to learning behavior characteristics of four types of the learning scenarios, a multi-constraint learning path recommendation model is proposed, in which the variables and their weighted coefficients considers different learning path preferences of the learners in different learning scenarios as well as learning resource organization and fragmented time. Thirdly, based on the proposed model and knowledge map, the design and implementation of a multi-constraint learning path recommendation algorithm is described. Finally, it is shown that the questionnaire results from over 110 e-learners verify the effectiveness of the proposed algorithm and show the similarity between the learners’ self-organized learning paths and the recommended learning paths. [ABSTRACT FROM AUTHOR]

In this paper, a novel probabilistic model of harmonic progressions and a generation scheme based on such model are presented. On the basis of the large amount of publications that show the stochastic nature of the music and the possibility of modelling it by means of statistical processes, this paper shows how to create a non-homogeneous Markov chain to automatically generate harmonic progressions by building a temporal reference of the internal beat structure of music to guide the progressions. Thus, this new model develops on the classic transition matrix to include a beat-dependent / temporal layer to model the residency time. The method for the automatic creation of harmonic progressions based on the model developed is presented after the model. The harmonic progressions generated by our scheme are coherent with the style of the training data employed and, thanks to the specific temporal layer designed, the musical mid-term and long-term dependencies that lead to a natural and logic cadence are taken into account. The model developed is usable for the automatic generation of harmonic patterns that can be used to enlarge the flexibility and creativity of pattern-based computational music composers. [ABSTRACT FROM AUTHOR]

Recently, Extreme learning machine (ELM), an efficient training algorithm for single-hidden-layer feedforward neural networks (SLFN), has gained increasing popularity in machine learning communities. In this paper the ELM based Area Under the ROC Curve (AUC) optimization algorithms are studied so as to further improve the performance of ELM for imbalanced datasets. For binary class problems, a novel ELM algorithm is proposed based on an efficient least square method. For multi-class problems, the following works are done in this paper: First of all, theoretical comparison analysis is proposed for the potential multi-class extensions of AUC; Secondly, a unified objective function for multi-class AUC optimization is proposed following the theoretical analysis; Subsequently, two ELM based multi-class AUC optimization algorithms called E L M M A U C and E L M m a c r o A U C respectively are proposed followed with complexity analyses; Finally, the generalization analysis is established for E L M M A U C in search of theoretical supports. Empirical study on a variety of real-world datasets show the effectiveness of our proposed algorithms. [ABSTRACT FROM AUTHOR]

This paper introduces a new hub-and-center transportation network problem for a new company competing against an operating company. The new company intends to locate p hubs and assign the center nodes to the located hubs in order to form origin–destination pairs. It desires not only to maximize the total captured flow in the market but also aims to minimize the total transportation cost. Three competition rules are established between the companies which must be abided. According to the competition rules, the new company can capture the full percentage of the traffic in each origin-destination pair if its transportation cost for each route is significantly less than of the competitor. If its transportation cost for each route is not significantly less than one of the competitors, only a certain percentage of the traffic can be captured. A bi-objective optimization model is proposed for the hub location problem on hand under a competitive environment. As the problem is shown to be NP-hard, a novel meta-heuristic algorithm called multi-objective biogeography-based optimization is developed. As there is no benchmark in the literature, a popular non-dominated sorting algorithm is utilized to validate the results obtained. Moreover, to enhance the performance of the proposed Pareto-based algorithms, this paper intends to develop a binary opposition-based learning as a diversity mechanism for both algorithms. The algorithms are tuned to solve the problem, based on which their performances are compared, ranked, and analyzed statistically. Finally, the applicability of the proposed approach and the solution methodologies are demonstrated in three steps. [ABSTRACT FROM AUTHOR]

Recently, there has been growing interest in social network analysis. Graph models for social network analysis are usually assumed to be a deterministic graph with fixed weights for its edges or nodes. As activities of users in online social networks are changed with time, however, this assumption is too restrictive because of uncertainty, unpredictability and the time-varying nature of such real networks. The existing network measures and network sampling algorithms for complex social networks are designed basically for deterministic binary graphs with fixed weights. This results in loss of much of the information about the behavior of the network contained in its time-varying edge weights of network, such that is not an appropriate measure or sample for unveiling the important natural properties of the original network embedded in the varying edge weights. In this paper, we suggest that using stochastic graphs, in which weights associated with the edges are random variables, can be a suitable model for complex social network. Once the network model is chosen to be stochastic graphs, every aspect of the network such as path, clique, spanning tree, network measures and sampling algorithms should be treated stochastically. In particular, the network measures should be reformulated and new network sampling algorithms must be designed to reflect the stochastic nature of the network. In this paper, we first define some network measures for stochastic graphs, and then we propose four sampling algorithms based on learning automata for stochastic graphs. In order to study the performance of the proposed sampling algorithms, several experiments are conducted on real and synthetic stochastic graphs. The performances of these algorithms are studied in terms of Kolmogorov-Smirnov D statistics, relative error, Kendall's rank correlation coefficient and relative cost. [ABSTRACT FROM AUTHOR]

In recent years, mainly due to the application of smartphones in this area, research in human activity recognition (HAR) has shown a continuous and steady growth. Thanks to its wide range of sensors, its size, its ease of use, its low price and its applicability in many other fields, it is a highly attractive option for researchers. However, the vast majority of studies carried out so far focus on laboratory settings, outside of a real-life environment. In this work, unlike in other papers, progress was sought on the latter point. To do so, a dataset already published for this purpose was used. This dataset was collected using the sensors of the smartphones of different individuals in their daily life, with almost total freedom. To exploit these data, numerous experiments were carried out with various machine learning techniques and each of them with different hyperparameters. These experiments proved that, in this case, tree-based models, such as Random Forest, outperform the rest. The final result shows an enormous improvement in the accuracy of the best model found to date for this purpose, from 74.39% to 92.97%. • This paper presents a comparison of the main machine learning algorithms for HAR. • A dataset taken in a real-life environment was used, unlike in other studies. • Experiments were done to get the best model configurations for long-themed activities. • This work also reviews the gyroscope's real influence to the final results. • The current approaches in this field, oriented towards real life, were highly improved. [ABSTRACT FROM AUTHOR]

Truth discovery is an effective way to eliminate data inconsistency by integrating different worker-provided values. Although directly conducting non-private truth discovery approaches based on uploaded noisy values after adding Laplace noise for continuous inputs guarantees rigorous local differential privacy (LDP), it may result in poor performance due to the lot of contained noise. First, the injected noise for privacy protection randomly sampled from Laplace distribution may be excessive even with a large privacy budget, as the above distribution is unbounded and drops sharply with respect to the x-axis. Built-in Gaussian noise also usually exists within these uploaded noisy values, which may also have a negative effect on the aggregated truths under LDP and makes the problem investigated in this paper far more challenging. In this paper, we focus on obtaining accurate truths in the above cases under rigorous LDP for continuous inputs, and present a novel solution TESLA. The key idea of this solution is that we let injected noise for privacy protection and inherent Gaussian noise only weakly negatively affect the weight estimation and true aggregation. In particular, we design a runtime filtering mechanism (RFM) to obtain the supremum and infimum for the values after adding Laplace noise by considering these two types of noise together. Moreover, we develop a probabilistic fusion mechanism (PFM) to get the fused values by adaptively using the obtained supremum and infimum. Furthermore, we devise a probabilistic weight mechanism (PWM) to obtain a more accurate weight for each worker. Therefore, truth discovery can be conducted based on the new weight of each worker and the filtered values. We provide theoretical analyses of TESLA 's utility, privacy and complexity. Experimental results demonstrate the effectiveness and efficiency of TESLA. We also extend and verify TESLA over typical mean estimation as well as standard deviation calculation, and various machine learning tasks (e.g., logistic regression, support vector machine (SVM) and neural network). Experimental results also demonstrate its superiority. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNNs) are the most important deep learning algorithms to classify images based on their visual features. CNNs architectures are made of convolution, pooling and fully-connected (FC) layers and their corresponding parameters, which affect classification performance significantly. Convolution parameters optimization for CNNs, referred as CPOCNN, is proposed in this paper. To the best of our knowledge, this is the first optimization model which assigns adaptive upper-bounds of convolution parameters depends on data dimension in current layer and number of remained layers to reach the output layer. For this task, a comprehensive mathematical model is presented for constant structures of CNNs and it is proven that more optimization space is explored in comparison with all state-of-the-art methods. In optimization process, number of convolution filters and type of pooling filters are selected randomly; dimension of pooling filters, zero-padding and stride are considered constant. CPOCNN has been evaluated on 7 publicly available datasets and compared with 53 competitive CNN models with constant and optimized structures. Interested results showed that CPOCNN not only outperforms state-of-the-art CNN methods but also enriches weak CNN models by improving their accuracies more than 35%. Source Code of this paper is available online at Github. 1 1 https://github.com/kohzadi/CHOCNN.git. [ABSTRACT FROM AUTHOR]

In multi-label learning, in order to improve the accuracy of classification, many scholars have considered the relationship between features and features, features and labels or labels and labels, but how to combine the correlation among them is rarely studied. Based on this, this paper proposes a multi-label learning algorithm with kernel extreme learning machine autoencoder. Firstly, the label space is reconstructed by using the non-equilibrium labels completion method in the label space. Then, the non-equilibrium labels space information is added to the input node of the kernel extreme learning machine autoencoder network, and the input features are output as the target. Finally, the kernel extreme learning machine is used for classification. Our method implements the information fusion between features and features, between labels and features, and between labels and labels. Compared with the traditional autoencoder network, the extreme learning machine autoencoder has no iterative process, which reduces the network training time and improves the classification accuracy. The experimental results of the proposed algorithm in the opening benchmark multi-label data sets show that the KELM-AE algorithm has some advantages over other comparative multi-label learning algorithms and the statistical hypothesis testing and stability analysis further illustrate the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

This paper presents an adaptive network intrusion detection (ANID) method based on the selective ensemble of kernel extreme learning machines (KELMs) with random features (termed ANID-SEoKELM), aiming at identifying various unauthorized uses, misuses and abuses of computer systems in real time. To generate a lightweight intrusion detector, multiple KELMs are learned independently based on the Bagging strategy with sparse random feature representation (SRFR), to reduce noise and redundant or irrelevant information in network connection instances and ensure the diversity of base learners for the effective ensemble of base learners. A marginal distance minimization (MDM)-based selective ensemble (MDMbSE) method is introduced to generate the ultimate intrusion detector. To ensure the adaptability of the intrusion detector, an incremental learning-based detection-model updating procedure is also derived. Extensive validation and comparative experiments on the benchmark KDD99 dataset and a hybrid heterogeneous network simulation platform mixed with wireless networks and Ethernet networks demonstrate that the ANID-SEoKELM is able to adapt to the dynamically changing network environments hence it can achieve higher detection accuracies stably and efficiently than classic single learner-based intrusion detection methods and representative ensemble-based intrusion detection methods. • A selective ensemble of KELMs-based intrusion detection method is proposed. • The KELM is introduced for the lightweight base classifier learning. • Random projection is adopted for the feature representation of network instances. • An incremental-learning-based online KELM updating approach is derived. • A margin distance minimization-based selective ensemble method is presented. [ABSTRACT FROM AUTHOR]

Ordinal regression is a special kind of machine learning problem, which aims to label patterns with an ordinal scale. Due to the ubiquitous existence of the ordering information in many practical cases, ordinal regression has received much attention and can be found in a great variety of applications. Meanwhile, Kernel Extreme Learning Machine (KELM), as the extension of Extreme Learning Machine in the framework of kernel learning, has shown its strength in many machine learning tasks. Nevertheless, existing KELM methods have paid little attention to ordinal regression especially in large-scale situations. In this paper, we consider the kernel technique and ordinal scales in labels at the same time, and propose a new KELM model for ordinal regression by exploiting a quadratic cost-sensitive encoding scheme. In order to make the training process more efficient in the big data scenario, a fast algorithm is designed based on the low rank approximation. The incomplete Cholesky factorization and the Sherman–Morrison–Woodbury formula are utilized together to avoid computing the inverse of the kernel matrix. The time complexity of the new algorithm is thus linear with the number of the training instances, which makes it suitable to the large-scale occasions. Numerical experiments on multiple public datasets validate the effectiveness and efficiency of the proposed methods. [ABSTRACT FROM AUTHOR]

Transfer learning handles a learning task on target domain by transferring the knowledge which was already learned from source domain, often under the situation where the labeled data in the target domain are insufficient or difficult to acquire. Transfer learning can be generally classified into two types, offline transfer learning and online transfer learning. As the former requires the data of target domain be given in advance, which may not be hold in real-life situations, the latter attracted more and more research in recent years. Online transfer learning deals with the situation where the data of target domain arrive in an online manner. Existing research on online transfer learning only deals with binary classification tasks, which appears to be important insufficiency. In this paper, the problem of online transfer learning for multi-class classification is studied, and an algorithm called Online Transfer Learning Algorithm for Multi-class Classification (OTLAMC) is proposed. OTLAMC learns a multi-class classifier in an online manner based on the knowledge from two sources, the obtained feedback of each datum in the target domain upon its arrival and the knowledge transferred from the source domain. A new loss function and a new updating method are proposed and adopted in OTLAMC, which improved the performance of OTLAMC on multi-class classification task. The mistake bound of OTLAMC is derived. Experiments on two widely used datasets illustrate that OTLAMC has good performance. • OTLAMC, an online transfer learning algorithm for multi-class classification, is proposed. • A new loss function and a new updating method are proposed and adopted in OTLAMC. • The mistake bound of the proposed OTLAMC is derived. • Experiments on two widely used datasets illustrate that OTLAMC has good performance. [ABSTRACT FROM AUTHOR]

Sparse reparameterization in Deep Neural Networks (DNNs) aims to achieve a better tradeoff between the network parameter count and performance. Recently, the lottery ticket hypothesis suppose that excellent sub-networks ("winning tickets") exist in dense randomly-initialized networks. These sparse sub-networks trained from scratch are able to reach the performance of their dense counterparts. Compared with Iterative Magnitude Pruning that relies on pruning strategies, the Continuous Sparsification algorithm learns the "winning tickets" with gradient-based methods, achieving better performance. In this paper, we propose Layer-wise Continuous Sparsification (LCS) scheme for finding sparse sub-networks, in which the parameterized relaxation of step functions used to remove network parameters in each layer is integrated into the DNN loss as an optimization objective. LCS utilizes a family of sigmoid functions to asynchronously filter important per-layer weights throughout training, yielding sparser and better sub-networks. Experiments show that our method surpasses state-of-the-art methods for sparse reparameterization. Additionally, the proposed method can be utilized as a regularization technique to further improve the accuracy of dense networks 1 1 Our code is publicly available at https://github.com/RiyaoDong/LCS.. [ABSTRACT FROM AUTHOR]

Nowadays, there is a growing interest among researchers in health monitoring using body sensor data due to its usage in diverse applications. Human Activity Recognition (HAR) utilizes smart wearable devices, particularly in healthcare, to help the elderly population improve their lifestyle by preventing forthcoming exceptional events such as health risks and falls. Although various types of sensors are available to recognize human behaviors, wearable sensors provide consistent information about the lifestyle and functionality of subjects. However, accurate recognition requires intelligent algorithms to extract temporal features of time series data from raw sensor readings. In order to recognize human activities more accurately from time series data, this paper proposes a crossover Aquila-optimized Tri-kernel Extreme Learning Machine (CSA-TkELM) classifier. The sensor data utilized to evaluate the effectiveness of the proposed CSA-TkELM classifier is obtained from the USC Human Activity Dataset (USC-HAD), the Public domain dataset for Real-life human activity recognition using smartphone Sensors (PDS-SS), and the OPPORTUNITY dataset. The direct application of raw time series data to the classifier affects its recognition performance. Therefore, preprocessing steps such as noise removal, handling missing values, standardization, and segmentation are executed. The proposed CSA-TkELM classifier extracts significant temporal features from the time series information and recognizes activities. Several performance metrics, such as recall, precision, accuracy, and F1 score, are measured to investigate the results of the proposed CSA-TkELM classifier. The proposed CSA-TkELM classifier achieves a recall of 98.1%, precision of 97.6%, accuracy of 97.9%, and F1 score of 97.3%, which are higher than those of other compared existing techniques. • Proposing CSA-TkELM to recognize human activity accurately from time series data. • Combining RBF, polynomial and linear with ELM to enhance classification performance. • Tuning TkELM parameters using CSA boosted Aquila optimizer. • The CSA-TkELM approach achieves greater recognition accuracy of 97.9%. [ABSTRACT FROM AUTHOR]

Phenotypes (i.e., symptoms and clinical signs) are essential for clinical diagnosis and research related to symptom science and precision health. As clinical observational manifestations of a disease, symptoms are clinically significant because they act as direct causes for patients to seek medical care and the primary indicators for clinicians to provide diagnosis/treatments. However, a comprehensive phenotypic knowledge base and high-quality symptom–gene associations are lacking. Therefore, a thorough understanding of the relationships between symptoms and other entities is urgently needed to support scientific research and clinical health care. In this paper, we constructed a systematic, large-scale, and high-quality symp tom- g ene a ssociations n etwork system named SympGAN (accessible at http://www.sympgan.org/). We provide access to the database with millions of associations between symptoms, genes, diseases, and drugs, as well as the system for users to search, analyze, knowledge inference, and present data visualization. We utilize state-of-the-art machine learning and deep learning algorithms as the backbone to form the final dataset. In addition, we utilize the RoBERTa-PubMed neural network for name entity recognition to assist in data screening. The knowledge graph is adopted to organize the relationships between different entities. We adopt ConvE, TuckER, and HypER methods for knowledge completion experiments to validate the quality of final knowledge graph triples. Based on the results, we provide online automatic knowledge inference interfaces. The system, SympGAN, has promising value for disease diagnosis, decision support in health care, precision health, and scientific research, as researchers and practitioners can easily access information about symptoms, diseases, targets, gene ontology, and drugs. [Display omitted] SympGAN is a comprehensive framework designed for the integration of symptom phenotypes, utilizing neural network embeddings and deep information extraction models. We have developed an integrative framework that establishes connections between symptoms and genes. This framework encompasses relationship inference through deep network embedding, literature mining using named entity recognition methods, and manual curation. Consequently, we have created a robust database and knowledge graph containing millions of associations between symptoms, genes, diseases, and drugs. SympGAN is readily accessible at http://www.sympgan.org/ , providing users with the ability to search, analyze, perform knowledge inference, and visualize information pertaining to these terminologies. The construction of SympGAN has successfully filled knowledge gaps and established millions of high-quality associations between symptoms, genotypes, diseases, and drugs. It holds tremendous potential for advancing precision health and symptom science. • We have developed SympGAN, a comprehensive, high-quality, and extensive knowledge graph-based system that encompasses the most comprehensive terminology set of 12,560 symptom phenotypes and their associations with genes, diseases, and drugs. • SympGAN has made a significant breakthrough by acquiring a comprehensive dataset for the knowledge graph, comprising 401,126 symptom–gene triples. These triples, along with their accompanying data, have undergone meticulous collection procedures to ensure exceptional quality. Our methodology involves employing the RoBERTa-PubMed model for named entity recognition (NER) and conducting literature mining from biomedical studies to gather pertinent information. Furthermore, we utilize sophisticated, high-precision algorithms to infer phenotypic associations with genes. • The website http://www.sympgan.org/ offers a comprehensive platform that enables users to conduct integrative searches and perform online knowledge inference and analysis. It serves as a centralized hub for exploring clinical knowledge associated with symptoms, as well as related diseases, genes, drugs, and molecular networks. This robust resource facilitates the interpretation and exploration of symptom phenotypes, particularly in understanding their genetic origins. By promoting precision health research and advancing the field of symptom science, http://www.sympgan.org/ significantly contributes to enhancing our understanding of symptoms and the underlying genetic factors involved. [ABSTRACT FROM AUTHOR]

From machine learning recommendation to deep learning recommendation, reinforcement learning recommendation, and recommendation model compression, the network structure of the recommendation algorithm becomes more complex, more expressive, and more lightweight with higher real-time performance. However, most of these models are dedicated to recommendations by mining ratings and text information, and researchers have optimized the network structures of many recommendation models. However, mining image information and enriching recommendation model features are less studied, and there is still some room for optimization of deep recommendation model structures and real-time performance. To this end, this paper proposes a model called MMDIN by adding multi-head and multi-modal modules to the DIN model. The multi-modal module extracts image features, which enriches the feature set that the model can use and strengthens the cross-combination and fitting expression capabilities of the model. Meanwhile, the multi-head mechanism extracts features from different dimensions, which further enhances the model prediction performance. Experimental results show that the proposed MMDIN model improves the recommendation prediction performance, and it outperforms state-of-the-art methods in comprehensive indicators. • Adding the Multimodal module allows the model to use richer features. • Add Multi-Head ResNet module to DIN to improve feature extraction capability. • Use spark for distributed in-memory computing to improve real-time performance. [ABSTRACT FROM AUTHOR]

Recommender systems, in real-world circumstances, tend to limit user exposure to certain topics and to overexpose them to others to maximize performance. However, repeated exposure to biased content could lead to the so-called echo chamber phenomenon: especially in social network environments, people encounter only information that reflects their previous beliefs and opinions, reinforcing them. This phenomenon could have worrying consequences for society, including the spread of aggressive, unhealthy, or risky behaviors. Some persons can be more affected than others by echo-chambers. We define as sensitive the users whose behavior could be influenced by the over- or under-exposure to certain items due to the echo-chamber effect, and as influential the items that could influence the behavior of such users. In this paper, we address the problem of recommending influential items to sensitive users. We formalize the problem and propose three techniques that can be used to diversify the distributions of influential items in order to positively affect sensitive users' behavior. Recommendations that meet this diversity criterion could potentially avoid dangerous societal consequences and simultaneously promote healthier lifestyles. We tested the proposed techniques in a real-world dataset by considering two different case studies that involved potentially aggressive and potentially depressed users. All techniques have been proven to be effective and allow high performance to be maintained while diversifying recommendations. [ABSTRACT FROM AUTHOR]

Machine learning algorithms need feature selection (FS) as a significant step towards filtering unnecessary data. This paper proposes a wrapper FS approach that combines the rat swarm optimization (RSO) algorithm with genetic operators to avoid local optimal. In the proposed approach the transfer functions (TFs) are added to balance local and global search by converting a continuous search space into a discrete space. Eight variants of the bmRSO algorithm were applied for classification purposes using a support vector machine (SVM) to increase accuracy and decrease the number of features over several chemical datasets. The eight bmRSO proposed methods and the original RSO were evaluated using the CEC'20 test suite and twelve datasets (eight chemical and four toxicity effect datasets) to verify their performance in complex optimization problems and FS over real datasets, respectively. Moreover, the binary versions of other stable metaheuristic algorithms such as Harris Hawks Optimization (HHO), Grey Wolf Optimization (GWO), Farmland Fertility Algorithm (FFA), Artificial Gorilla Troops Optimizer (GTO), African Vultures Optimization Algorithm (AVOA), Runge Kutta Optimizer's (RUN), and Slime Mould Algorithm (SMA) were used to compare the results obtained by the best variant of the bmRSO. Eventually, the experimental results have revealed that in most of the tests, the proposed bmRSO1 has achieved efficient search results with higher convergence speeds without increasing additional computational efforts. From the twelve datasets, the MAO dataset reached the highest results compared with other datasets, so the proposed method, bmRSO1-SVM, achieved an accuracy of 98.201% and a 20.001 number of selected features. • An efficient bmRSO algorithm is proposed based on transfer functions. • CEC'20 test suite is utilized for verification of bmRSO performance. • bmRSO is proposed for cheminformatics classification tasks. • bmRSO is analyzed using various analysis metrics. • The performance of the bmRSO is better than other competitor algorithms. [ABSTRACT FROM AUTHOR]

Learning class-imbalance data has become a challenging task in machine learning. Oversampling is an effective way to achieve rebalancing between classes by generating new minority instances. However, most existing oversampling methods exhibit significant ill-posedness due to the involvement of k -nearest neighbors. Recently, several novel techniques have been developed using natural neighbors instead of k -nearest neighbors. But these budding techniques fail to take into account the natural distribution of data, resulting in degraded adaptability of natural neighbors to data characteristics. In this paper, we propose an adaptive and completely parameter-free synthetic oversampling method called NanBDOS (short for borderline oversampling via natural neighbor search). NanBDOS first determines the intrinsic neighbor relationship between instances through a natural neighbor search procedure. Then, those informative borderline instances are identified and used as base instances, i.e., sampling seeds. NanBDOS assigns dynamic sampling weights to the base instances so that the data complexity can be well represented. Finally, new synthetic instances are generated by interpolating between the base instances and their natural neighbors. NanBDOS is experimentally compared with seven baseline methods on twenty-four real-world datasets. The results confirm the effectiveness of the proposed method. Moreover, the statistical analysis also indicates its higher-level Friedman ranking. [ABSTRACT FROM AUTHOR]

Data quality is an essential element necessary for the development of a successful machine-learning project. One of the biggest challenges in various real-world application domains is class imbalance. This paper proposes a new framework for oversampling credit data by combining two deep learning techniques: autoencoders and generative adversarial networks. A trivial autoencoder (TAE) is used to change data representation, and modified generative adversarial networks (GAN) are used to create new instances from random noise. The experiment on three different datasets demonstrates that the same classifier achieves a better area under the receiver operating characteristic curve (AUC) on datasets augmented by the proposed framework compared to datasets oversampled by other techniques. Additionally, the results show that datasets balanced by the new framework influence the classifier to change the prediction error types, significantly reducing false negatives; more expensive misclassification case in the imbalance learning. The improvements are significant, and considering the change in error distribution, the proposed technique is an excellent complement to existing oversampling techniques. • Credit risk datasets are imbalanced and contain discrete and continuous variables. • A new framework for oversampling tabular data is proposed. • Deep learning techniques for generating new data are applied to balance credit datasets. • The classifier performs better on AUC and recall after adding new instances to the dataset. [ABSTRACT FROM AUTHOR]

The COVID-19 pandemic has resulted in a surge of fake news, creating public health risks. However, developing an effective way to detect such news is challenging, especially when published news involves mixing true and false information. Detecting COVID-19 fake news has become a critical task in the field of natural language processing (NLP). This paper explores the effectiveness of several machine learning algorithms and fine-tuning pre-trained transformer-based models, including Bidirectional Encoder Representations from Transformers (BERT) and COVID-Twitter-BERT (CT-BERT), for COVID-19 fake news detection. We evaluate the performance of different downstream neural network structures, such as CNN and BiGRU layers, added on top of BERT and CT-BERT with frozen or unfrozen parameters. Our experiments on a real-world COVID-19 fake news dataset demonstrate that incorporating BiGRU on top of the CT-BERT model achieves outstanding performance, with a state-of-the-art F1 score of 98%. These results have significant implications for mitigating the spread of COVID-19 misinformation and highlight the potential of advanced machine learning models for fake news detection. [ABSTRACT FROM AUTHOR]

As one of the classical machine learning algorithms, twin support vector machine (TWSVM) can construct two nonparallel hyperplanes, which keeps the hyperplane close to points of one class, but far away from the other class. Being different from the classical SVM algorithm, TWSVM can reduce the computational complexity by replacing one large quadratic programming problems (QPPs) with a pair of smaller ones. Although compared to SVM classifier, TWSVM is four times faster, it is still insufficient in resisting noise or outliers. In this paper, we propose a twin depth support vector machine (TDSVM), a binary SVM classifier that considers the influence of depth when calculating the distance. A novel average depth is proposed and applied on TWSVM to construct a robust SVM framework, which can identify outliers in the dataset. By strengthening the center and weakening the edge, a better generalization performance is achieved, and the SRM principle is also implemented. TDSVM can be applied to any place where TWSVM can be applied, which is also useful for reducing the influence of outliers or noise on data and obtaining more robust results. Finally, the advantages of the method are verified theoretically and empirically by experiments. Experimental results on eight UCI datasets and one synthetic dataset demonstrate the effectiveness and robustness of TDSVM. The classification accuracy of TDSVM is better than other compared algorithms on almost every dataset, whether a certain percentage of Gaussian noise is introduced or not. • A novel average depth is proposed to identify outliers in the dataset. • The average depth is innovatively applied to TWSVM to obtain TDSVM. • The proposed TDSVM is extended from linear classifiers to nonlinear classifiers. • TDSVM can significantly address the impact of noise or outliers. [ABSTRACT FROM AUTHOR]

Gradient boosted trees (GBTs) are ubiquitous models used by researchers, machine learning (ML) practitioners, and data scientists because of their robust performance, interpretable behavior, and ease-of-use. One critical challenge in training GBTs is the tuning of their hyperparameters. In practice, selecting these hyperparameters is often done manually. Recently, the ML community has advocated for tuning hyperparameters through black-box optimization and developed state-of-the-art systems to do so. However, applying such systems to tune GBTs suffers from two drawbacks. First, these systems are not model-aware , rather they are designed to apply to a generic model; this leaves significant optimization performance on the table. Second, using these systems requires domain knowledge such as the choice of hyperparameter search space, which is an antithesis to the automatic experimentation that black-box optimization aims to provide. In this paper, we present SigOpt Mulch, a model-aware hyperparameter tuning system specifically designed for automated tuning of GBTs that provides two improvements over existing systems. First, Mulch leverages powerful techniques in metalearning and multifidelity optimization to perform model-aware hyperparameter optimization. Second, it automates the process of learning performant hyperparameters by making intelligent decisions about the optimization search space, thus reducing the need for user domain knowledge. These innovations allow Mulch to identify good GBT hyperparameters far more efficiently—and in a more seamless and user-friendly way—than existing black-box hyperparameter tuning systems. • Mulch is a system developed to automate hyperparameter optimization of GBTs. • Mulch chooses the HPO search space, reducing the need for domain expertise. • Mulch uses metalearning and multifidelity optimization to accelerate performance. • Mulch delivers superior performance compared to black-box software like Optuna. [ABSTRACT FROM AUTHOR]

Constrained optimization (CO) has made a profound impact in solving many real-world problems. Due to the high computation burden in exact solvers, data-driven CO based on machine learning techniques is recently receiving extensive research interests for its capability to solve CO problems in real time. The existing data-driven CO approaches only serve for optimization problems with rather simple constraints that can be directly incorporated into model training. However, constraints that are computationally infeasible or burdensome to evaluate are commonly experienced in realistic optimization applications, especially in the engineering sector. This paper proposes an athlete–referee dual learning system (ARDLS) for end-to-end CO with large-scale complex constraints, where an athlete model is trained as the main optimizer while a referee model is trained as a probabilistic constraint classifier to guide the athlete training. A risk-based constrained loss function is designed to fine-tune the athlete model for constraint satisfaction. A case study on electric power system emergency control application is conducted to validate the proposed ARDLS, where the testing results demonstrate the excellent capability of ARDLS to improve the likelihood of satisfying large-scale complex constraints in CO. [ABSTRACT FROM AUTHOR]

Humans perform Bayesian inference in a wide variety of tasks, which can help people make selection decisions effectively and therefore enhances learning efficiency and accuracy. Inspired by this fact, this paper presents a novel human learning optimization algorithm with Bayesian inference learning (HLOBIL), in which a Bayesian inference learning operator (BILO) is developed to utilize the inference strategy for enhancing learning efficiency. The in-depth analysis shows that the proposed BILO can efficiently improve the exploitation ability of the algorithm as it can achieve the optimal values and retrieve the optimal information with the accumulated search information. Besides, the exploration ability of HLOBIL is also strengthened by the inborn characteristics of Bayesian inference. The experimental results demonstrate that the developed HLOBIL is superior to previous HLO variants and other state-of-art algorithms with its improved exploitation and exploration abilities. [ABSTRACT FROM AUTHOR]

A critical aspect of dimensionality reduction is to assess the quality of selected (or produced) feature subsets properly. Feature subset assessment in machine learning refers to split a given feature subset into a training set, which is used to estimate the parameters of a classification model, and a test set used to estimate the predictive performance of the model. Then, averaging the results of multiple splitting (i.e., Cross-Validation, CV) is commonly used to decrease the variance of the estimator. But in practice, CV scheme is very computationally expensive. In this paper, we propose a new statistics index method called LW-index for evaluation of feature subset and dimensionality reduction algorithms in general. The proposed method is a type of “classical statistics” approach that uses the feature subset to compute an empirical estimate of the quality of feature subset. A large number of performance comparisons with the machine learning approach conducted on fourteen benchmark collections show that the proposed LW-index is highly correlated with the external indices (i.e., MacroF 1 , MicroF 1 ) of SVM and Centroid-Based Classifier (CBC) trained by five-fold CV scheme. Furthermore, the experimental results indicate that LW-index has the same performance as the traditional CV scheme for evaluating the dimensionality reduction algorithms and it is more efficient than the traditional methodology. Therefore, one contribution of this paper is to present an alternative methodology, based on an internal index typically used in the unsupervised learning context, that is computationally cheaper than the traditional CV methodology. Another contribution is to propose a new internal index that behaves better than other similar indices widely used in clustering and shows high correlation with the results obtained by the traditional methodology. [ABSTRACT FROM AUTHOR]

This paper considers a parallel algorithm for Bayesian network structure learning from large data sets. The parallel algorithm is a variant of the well known PC algorithm. The PC algorithm is a constraint-based algorithm consisting of five steps where the first step is to perform a set of (conditional) independence tests while the remaining four steps relate to identifying the structure of the Bayesian network using the results of the (conditional) independence tests. In this paper, we describe a new approach to parallelization of the (conditional) independence testing as experiments illustrate that this is by far the most time consuming step. The proposed parallel PC algorithm is evaluated on data sets generated at random from five different real-world Bayesian networks. The algorithm is also compared empirically with a process-based approach where each process manages a subset of the data over all the variables on the Bayesian network. The results demonstrate that significant time performance improvements are possible using both approaches. [ABSTRACT FROM AUTHOR]

Differently from Vector-pattern-oriented Classifier Design (VecCD), Matrix-pattern-oriented Classifier Design (MatCD) is expected to manipulate matrix-oriented patterns directly rather than turning them into a vector, and further demonstrated its effectiveness. However, some prior information, such as the local sensitive discriminant information among matrix-oriented patterns, might be neglected by MatCD. To overcome such flaw, a new regularization term named R LSD is adopted into MatCD by taking advantage of Locality Sensitive Discriminant Analysis (LSDA) in this paper. In detail, the objective function of LSDA is modified and transformed into the regularization term R LSD to explore the local sensitive discriminant information among matrix-oriented patterns. In the implementation, R LSD is collaborated with one typical MatCD, whose name is Matrix-pattern-oriented Ho-Kashyap Classifier (MatMHKS), so as to create a new classifier based on local sensitive discriminant information named LSDMatMHKS for short. Finally, comprehensive experiments are designed to validate the effectiveness of LSDMatMHKS. The major contributions of this paper can be concluded as (1) improving the classification performance and the learning ability of MatCD, (2) introducing local sensitive discriminant information into MatCD and extending the application scenario of LSDA, and (3) validating and analyzing the feasibility and effectiveness of R LSD . [ABSTRACT FROM AUTHOR]

When a new book is launched the publisher faces the problem of how many books should be printed for delivery to bookstores; printing too many is the main issue, since it implies a loss of investment due to inventory excess, but printing too few will also have a negative economic impact. In this paper, we are tackling the problem of predicting total sales in order to print the right amount of books and doing so even before the book has reached the stores. A real dataset including the complete sales data for books published in Spain across several years has been used. We have conducted an analysis in three stages: an initial exploratory analysis, by means of data visualisation techniques; a feature selection process, using different techniques to find out what are the variables that have more impact on sales; and a regression or prediction stage, in which a set of machine learning methods has been applied to create forecasting models for book sales. The obtained models are able to predict sales from pre-publication data with remarkable accuracy, and can be visualised as simple decision trees. Thus, these can be used as decision-aid tools for publishers, which can provide a reliable guidance on the decision process of publishing a book. This is also shown in the paper by addressing four example cases of representative publishers, regarding their number of sales and the number of different books they sell. [ABSTRACT FROM AUTHOR]

This paper presents HBIN-LBD, a novel literature-based discovery (LBD) method that exploits the lexico-citation structures within the heterogeneous bibliographic information network (HBIN) graphs. Unlike other existing LBD methods, HBIN-LBD harnesses the metapath features found in HBIN graphs for discovering the latent associations between scientific papers published in otherwise disconnected research areas. Further, this paper investigates the effects of incorporating semantic and topic modeling components into the proposed models. Using time-sliced historical bibliographic data, we demonstrate the performance of our method by reconstructing two LBD hypotheses: the Fish Oil and Raynaud’s Syndrome hypothesis and the Migraine and Magnesium hypothesis. The proposed method is capable of predicting the future co-citation links between research papers of these previously disconnected research areas with up to 88.86% accuracy and 0.89 F-measure. [ABSTRACT FROM AUTHOR]

Trust can help participants in online social communities to make decisions; however, it is a challenge for systems to map trust into computational models because of its subjective properties. Also, many online social communities are sparsely connected. Therefore, it is necessary to introduce mechanisms which can infer indirect trust among participants who are not directly connected. We provide a survey of existing trust management systems for online social communities. We also list four types of attacks, and analyze existing systems’ vulnerabilities. Compared with previous surveys, our survey takes trust modeling, trust inference, and attacks into account. Although there are several survey papers about global trust/reputation related attacks, the main contribution of this paper is that we consider trust inference and potential local trust related attacks. [ABSTRACT FROM AUTHOR]

In this paper we propose a new approach to learning fuzzy cognitive maps (FCMs) as a predictive model for time series forecasting. The first contribution of this paper is the dynamic optimization of the FCM structure, i.e., we propose to select concepts involved in the FCM model before every prediction is made. In addition, the FCM transformation function together with the corresponding parameters are proposed to be optimized dynamically. Finally, the FCM weights are learned. In this way, the entire FCM model is learned in a completely new manner, i.e., it is continuously adapted to the current local characteristics of the forecasted time series. To optimize all of the aforementioned elements, we apply and compare 5 different population-based algorithms: genetic, particle swarm optimization, simulated annealing, artificial bee colony and differential evolution. For the evaluation of the proposed approach we use 11 publicly available data sets. The results of comparative experiments provide evidence that our approach offers a competitive forecasting method that outperforms many state-of-the-art forecasting models. We recommend to use our FCM-based approach for the forecasting of time series that are linear and tend to be trend stationary. [ABSTRACT FROM AUTHOR]