Instances Of Class Research Articles

IMPLEMENTATION OF RANDOM FOREST AND SMOTE METHODS FOR ECONOMIC STATUS CLASSIFICATION IN CIREBON CITY

The Indonesian government has tried various methods to eradicate poverty throughout the country, one of which is the fair distribution of social assistance to households based on their economic status classification. The determination of social assistance recipients can be influenced by political factors or personal relationships and can be misused, leading to assistance being given to individuals with specific political connections or support. This research aims to develop a household economic status classification system in Cirebon City using the Random Forest algorithm to address these issues. The data used in this research experienced an imbalance in the number of class instances, where the high-class instances were much fewer than the low and medium classes. To address this, the Synthetic Minority Oversampling Technique (SMOTE) was employed. In this study, various testing scenarios were conducted to obtain the best model for accurately predicting household economic status. Based on the research results, the best testing was achieved using Random Forest and SMOTE with a random state of 0, obtaining an accuracy of 93% and excellent performance in classifying each class. When testing unlabeled data, Random Forest successfully predicted 24 out of 30 actual data, resulting in an accuracy rate of 80%. Although this accuracy is lower than what was achieved by Random Forest and SMOTE with a random state of 0, it can be said that the application successfully classifies household economic status in Cirebon City effectively.

A metric learning-based method for biomedical entity linking.

Biomedical entity linking task is the task of mapping mention(s) that occur in a particular textual context to a unique concept or entity in a knowledge base, e.g., the Unified Medical Language System (UMLS). One of the most challenging aspects of the entity linking task is the ambiguity of mentions, i.e., (1) mentions whose surface forms are very similar, but which map to different entities in different contexts, and (2) entities that can be expressed using diverse types of mentions. Recent studies have used BERT-based encoders to encode mentions and entities into distinguishable representations such that their similarity can be measured using distance metrics. However, most real-world biomedical datasets suffer from severe imbalance, i.e., some classes have many instances while others appear only once or are completely absent from the training data. A common way to address this issue is to down-sample the dataset, i.e., to reduce the number instances of the majority classes to make the dataset more balanced. In the context of entity linking, down-sampling reduces the ability of the model to comprehensively learn the representations of mentions in different contexts, which is very important. To tackle this issue, we propose a metric-based learning method that treats a given entity and its mentions as a whole, regardless of the number of mentions in the training set. Specifically, our method uses a triplet loss-based function in conjunction with a clustering technique to learn the representation of mentions and entities. Through evaluations on two challenging biomedical datasets, i.e., MedMentions and BC5CDR, we show that our proposed method is able to address the issue of imbalanced data and to perform competitively with other state-of-the-art models. Moreover, our method significantly reduces computational cost in both training and inference steps. Our source code is publicly available here.

Instance space analysis for 2D bin packing mathematical models

In this paper, we apply Instance Space Analysis (ISA) to study the two-dimensional bin-packing problem. We consider classical and newly-generated instances to test the performance of four mixed-integer programming (MIP) models from the literature. This is the first time ISA is used to compare MIP models. We set as a performance metric the time taken by the black-box MIP solver CPLEX to obtain a proven optimal solution when running each model. Our results provide a new perspective on the different models’ performance according to each instance’s features.

Solving imbalanced learning with outlier detection and features reduction

A critical problem for several real world applications is class imbalance. Indeed, in contexts like fraud detection or medical diagnostics, standard machine learning models fail because they are designed to handle balanced class distributions. Existing solutions typically increase the rare class instances by generating synthetic records to achieve a balanced class distribution. However, these procedures generate not plausible data and tend to create unnecessary noise. We propose a change of perspective where instead of relying on resampling techniques, we depend on unsupervised features engineering approaches to represent records with a combination of features that will help the classifier capturing the differences among classes, even in presence of imbalanced data. Thus, we combine a large array of outlier detection, features projection, and features selection approaches to augment the expressiveness of the dataset population. We show the effectiveness of our proposal in a deep and wide set of benchmarking experiments as well as in real case studies.

A secure solution based on load-balancing algorithms between regions in the cloud environment

The problem treated in this article is the storage of sensitive data in the cloud environment and how to choose regions and zones to minimize the number of transfer file events. Handling sensitive data in the global internet network many times can increase risks and minimize security levels. Our work consists of scheduling several files on the different regions based on the security and load balancing parameters in the cloud. Each file is characterized by its size. If data is misplaced from the start it will require a transfer from one region to another and sometimes from one area to another. The objective is to find a schedule that assigns these files to the appropriate region ensuring the load balancing executed in each region to guarantee the minimum number of migrations. This problem is NP-hard. A novel model regarding the regional security and load balancing of files in the cloud environment is proposed in this article. This model is based on the component called “Scheduler” which utilizes the proposed algorithms to solve the problem. This model is a secure solution to guarantee an efficient dispersion of the stored files to avoid the most storage in one region. Consequently, damage to this region does not cause a loss of big data. In addition, a novel method called the “Grouping method” is proposed. Several variants of the application of this method are utilized to propose novel algorithms for solving the studied problem. Initially, seven algorithms are proposed in this article. The experimental results show that there is no dominance between these algorithms. Therefore, three combinations of these seven algorithms generate three other algorithms with better results. Based on the dominance rule, only six algorithms are selected to discuss the performance of the proposed algorithms. Four classes of instances are generated to measure and test the performance of algorithms. In total, 1,360 instances are tested. Three metrics are used to assess the algorithms and make a comparison between them. The experimental results show that the best algorithm is the “Best-value of four algorithms” in 86.5% of cases with an average gap of 0.021 and an average running time of 0.0018 s.

Category Generalization After Entrenched Versus Probabilistic Erroneous Feedback

Abstract The present study investigated the impact of occasional erroneous feedback on category learning. Subjects were presented with exemplars from multiple prototype categories, where the exemplars were novel on each learning block (nonrepeat) or were continuously recycled through learning (repeat). In training, selected training instances were associated with a label different from other members drawn from the same prototype category. After learning, subjects received a common transfer test requiring either the classification of novel instances (Experiment 1) or the discrimination of old from new instances (Experiments 2 and 3). The major results were that as expected, learning was more accurate in the repeat condition, with subjects learning to accurately classify both the correct and incorrect feedback patterns by the terminal learning block; learning in the nonrepeat condition was worse, with subjects misclassifying the erroneous feedback patterns at an increasing rate through learning; nonetheless, classification and recognition transfer were somewhat higher after nonrepeat learning; and subjects readily discriminated old from new after repetition training but not after nonrepetition training. Overall, the enhanced classification and recognition after nonrepeat training mirrored our results reported in a similar paradigm that used only correct feedback. Formal modeling suggested that a pure prototype model could not predict learning or transfer after repetition training, whereas a prototype model was superior at capturing results after nonrepetition learning.

Parameter-free basis allocation for efficient multiple metric learning

Metric learning involves learning a metric function for distance measurement, which plays an important role in improving the performance of classification or similarity-based algorithms. Multiple metric learning is essential for efficiently reflecting the local properties between instances, as single metric learning has limitations in reflecting the nonlinear structure of complex datasets. Previous research has proposed a method for learning a smooth metric matrix function through data manifold to address the challenge of independently learning multiple metrics. However, this method uses the basic distance-based clustering algorithm to set the anchor points, which are the basis for local metric learning, and the number of basis metrics is dependent on the user. We propose a new method that can assign sophisticated anchor points by iteratively partitioning to identify mixed clusters of multi-class instances and cluster the most similar class instances together. In an experiment, we demonstrate the reliability of the automatically set parameter by comparison with the distribution of error rates according to the number of basis metrics of the existing algorithm. Furthermore, we show the superior performance of the proposed method over a fixed parameter setting of existing algorithms and confirm the relative classification accuracy superiority through performance comparison with baseline algorithms.

More Reliable Neighborhood Contrastive Learning for Novel Class Discovery in Sensor-Based Human Activity Recognition

Human Activity Recognition (HAR) systems have made significant progress in recognizing and classifying human activities using sensor data from a variety of sensors. Nevertheless, they have struggled to automatically discover novel activity classes within massive amounts of unlabeled sensor data without external supervision. This restricts their ability to classify new activities of unlabeled sensor data in real-world deployments where fully supervised settings are not applicable. To address this limitation, this paper presents the Novel Class Discovery (NCD) problem, which aims to classify new class activities of unlabeled sensor data by fully utilizing existing activities of labeled data. To address this problem, we propose a new end-to-end framework called More Reliable Neighborhood Contrastive Learning (MRNCL), which is a variant of the Neighborhood Contrastive Learning (NCL) framework commonly used in visual domain. Compared to NCL, our proposed MRNCL framework is more lightweight and introduces an effective similarity measure that can find more reliable k-nearest neighbors of an unlabeled query sample in the embedding space. These neighbors contribute to contrastive learning to facilitate the model. Extensive experiments on three public sensor datasets demonstrate that the proposed model outperforms existing methods in the NCD task in sensor-based HAR, as indicated by the fact that our model performs better in clustering performance of new activity class instances.

InsPLAD: A Dataset and Benchmark for Power Line Asset Inspection in UAV Images

ABSTRACT Power line maintenance and inspection are essential to avoid power supply interruptions, reducing its high social and financial impacts yearly. Automating power line visual inspections remains a relevant open problem for the industry due to the lack of public real-world datasets of power line components and their various defects to foster new research. This paper introduces InsPLAD, a Power Line Asset Inspection Dataset and Benchmark containing 10,607 high-resolution Unmanned Aerial Vehicles colour images. The dataset contains 17 unique power line assets captured from real-world operating power lines. Additionally, five of those assets present six defects: four of which are corrosion, one is a broken component, and one is a bird’s nest presence. All assets were labelled according to their condition, whether normal or the defect name found on an image level. We thoroughly evaluate state-of-the-art and popular methods for three image-level computer vision tasks covered by InsPLAD: object detection, through the AP metric; defect classification, through Balanced Accuracy; and anomaly detection, through the AUROC metric. InsPLAD offers various vision challenges from uncontrolled environments, such as multi-scale objects, multi-size class instances, multiple objects per image, intra-class variation, cluttered background, distinct point-of-views, perspective distortion, occlusion, and varied lighting conditions. To the best of our knowledge, InsPLAD is the first large real-world dataset and benchmark for power line asset inspection with multiple components and defects for various computer vision tasks, with a potential impact to improve state-of-the-art methods in the field. It will be publicly available in its integrity on a repository with a thorough description. It can be found at https://github.com/andreluizbvs/InsPLAD/.

Diabetes classification using MapReduce-based capsule network

Big data analytics is a complex exploratory process to uncover hidden data information from vast collections of data. It often provides enormous information from diverse sources and the use of analytics provides confined knowledge from the collected noisy data. In the case of diabetes data, there exist a massive collection of patient data that relates to significant information on patient health and its critical nature. In order of validating and analysing the data to get desired information about a patient and their health risk from the vast collection of data, the study uses bigdata based deep learning analytics. This study uses a Deep Learning Model namely capsule network (CapsNet) is executed on a MapReduce framework. The CapsNet present in the MapReduce framework enables the classification of instances via proper regulations. This model after suitable training with the training dataset enables optimal classification of instances to detect the nature of the risk of a patient. The validation conducted on the test dataset shows that the proposed CapsNets-based MapReduce model obtains increased accuracy, recall, and F-score than the conventional MapReduce and deep learning models.

GMMSampling: a new model-based, data difficulty-driven resampling method for multi-class imbalanced data

Learning from multi-class imbalanced data has still received limited research attention. Most of the proposed methods focus on the global class imbalance ratio only. In contrast, experimental studies demonstrated that the imbalance ratio itself is not the main difficulty in the imbalanced learning. It is the combination of the imbalance ratio with other data difficulty factors, such as class overlapping or minority class decomposition into various subconcepts, that significantly affects the classification performance. This paper presents GMMSampling—a new resampling method that exploits information about data difficulty factors to clear class overlapping regions from majority class instances and to simultaneously oversample each subconcept of the minority class. The experimental evaluation demonstrated that the proposed method achieves better results in terms of G-mean, balanced accuracy, macro-AP, MCC and F-score than other related methods.

Self-adaptive General Variable Neighborhood Search algorithm for parallel machine scheduling with unrelated servers

This paper introduces a machine learning-augmented General Variable Neighborhood Search algorithm, termed Self-adaptive General Variable Neighborhood Search (SGVNS), for solving the two-identical parallel machine scheduling problem with unrelated servers. The studied problem is of high practical relevance in various industries and presents a significant gap in the existing literature. The motivation for this study stems from the healthcare industry, specifically surgery scheduling with limited anesthesiologists’ availability. Two mathematical formulations are presented to solve this scheduling problem. However, these formulations are limited to small-scale problems, prompting the need for an efficient metaheuristic. Therefore, a novel variant of the General Variable Neighborhood Search metaheuristic is developed, incorporating a self-adaption mechanism based on an Extreme Learning Machine (ELM) algorithm. Extensive experiments are conducted to assess the performance of the SGVNS metaheuristic on various classes of instances. The results demonstrate the effectiveness and efficiency of SGVNS in solving previously unsolved small-sized instances, outperforming both formulations while maintaining reasonable computation times.

Learning from class-imbalanced data using misclassification-focusing generative adversarial networks

This paper presents a novel end-to-end oversampling-classification approach, which we refer to as imbalanced data-classifying generative adversarial network (ImbGAN), for imbalanced data classification. ImbGAN has a classifier-embedded structure within a GAN and consists of five components: (1) generator, (2) discriminator, (3) classifier, (4) storage for misclassified minority class data, and (5) storage for artificial minority class data. By iterative interaction with the embedded classifier, the first two components generate artificial minority class instances that are similar to minority class instances misclassified by the classifier. Therefore, these three networks are iteratively and simultaneously trained. The misclassified and artificial minority class instances are stored in the fourth and fifth components, respectively. These two components are also updated as iterations proceed. Our method obtains the final classification model from a single learning process, while most artificial data generation methods for imbalanced data classification go through an additional process for training classifiers after artificial data generation. Numerical experiments based on tabular, image, and text datasets confirm that the proposed method outperforms well-known synthetic sampling methods.

A Diversity-Based Synthetic Oversampling Using Clustering for Handling Extreme Imbalance

Imbalanced data are typically observed in many real-life classification problems. However, mainstream machine learning algorithms are mostly designed with the underlying assumption of a relatively well-balanced distribution of classes. The mismatch between reality and algorithm assumption results in a deterioration of classification performance. One form of approach to address this problem is through re-sampling methods, although its effectiveness is limited; most re-sampling methods fail to consider the distribution of minority and majority instances and the diversity within synthetically generated data. Diversity becomes increasingly important when minority data becomes more sparse, as each data point becomes more valuable. They should all be considered during the generation process instead of being regarded as noise. In this paper, we propose a cluster-based diversity re-sampling method, combined with NOAH algorithm. Neighbourhood-based Clustering Diversity Over-sampling (NBCDO) is introduced with the aim to complement our previous cluster-based diversity algorithm Density-based Clustering Diversity Over-sampling (DBCDO). It first uses a neighbourhood-based clustering algorithm to consider the distribution of both minority and majority class instances, before applying NOAH algorithm to encourage diversity optimisation during the generation of synthetic instances. We demonstrate the implementation of both cluster-based diversity methods by conducting experiments over 10 real-life datasets with ≤ 5% imbalance ratio and show that our proposed cluster-based diversity algorithm (NBCDO, DBCDO) brings performance improvements over its comparable methods (DB-SMOTE, MAHAKIL, KMEANS-SMOTE, MC-SMOTE).

Scheduling algorithms for data-protection based on security-classification constraints to data-dissemination

Communication networks have played a vital role in changing people’s life. However, the rapid advancement in digital technologies has presented many drawbacks of the current inter-networking technology. Data leakages severely threaten information privacy and security and can jeopardize individual and public life. This research investigates the creation of a private network model that can decrease the number of data leakages. A two-router private network model is designed. This model uses two routers to manage the classification level of the transmitting network packets. In addition, various algorithmic techniques are proposed. These techniques solve a scheduling problem. This problem is to schedule packets through routers under a security classification level constraint. This constraint is the non-permission of the transmission of two packets that belongs to the same security classification level. These techniques are the dispatching rule and grouping method. The studied problem is an NP-hard. Eight algorithms are proposed to minimize the total transmission time. A comparison between the proposed algorithms and those in the literature is discussed to show the performance of the proposed scheme through experimentation. Four classes of instances are generated. For these classes, the experimental results show that the best-proposed algorithm is the best-classification groups’ algorithm in 89.1% of cases and an average gap of 0.001. In addition, a benchmark of instances is used based on a real dataset. This real dataset shows that the best-proposed algorithm is the best-classification groups’ algorithm in 88.6% of cases and an average gap of less than 0.001.

KNNENS: A k-Nearest Neighbor Ensemble-Based Method for Incremental Learning Under Data Stream With Emerging New Classes.

In this brief, we investigate the problem of incremental learning under data stream with emerging new classes (SENC). In the literature, existing approaches encounter the following problems: 1) yielding high false positive for the new class; i) having long prediction time; and 3) having access to true labels for all instances, which is unrealistic and unacceptable in real-life streaming tasks. Therefore, we propose the k -Nearest Neighbor ENSemble-based method (KNNENS) to handle these problems. The KNNENS is effective to detect the new class and maintains high classification performance for known classes. It is also efficient in terms of run time and does not require true labels of new class instances for model update, which is desired in real-life streaming classification tasks. Experimental results show that the KNNENS achieves the best performance on four benchmark datasets and three real-world data streams in terms of accuracy and F1-measure and has a relatively fast run time compared to four reference methods. Codes are available at https://github.com/Ntriver/KNNENS.

An ensemble learning approach for anomaly detection in credit card data with imbalanced and overlapped classes

Electronic payment methods have become increasingly popular for business transactions, both online and in-person, across the globe. Anomalies like online fraud and default payments, which can result in substantial financial losses, have become more common as the usage of credit cards in online purchases has increased. To address this issue, researchers have explored various machine learning models and their ensemble techniques for detecting anomalies in credit card transaction data. However, detecting anomalies in this data can be challenging due to overlapping class samples and an imbalanced class distribution. Therefore, the detection rate of anomalies from minority class samples is relatively low, and general learning algorithms can be biased towards the majority class samples. In this paper, we propose a model called Credit Card Anomaly Detection (CCAD) that leverages the base learners paradigm and meta-learning ensemble techniques to improve the detection rate of credit card anomalies. We utilize four outlier detection algorithms as base learners and XGBoost algorithm as meta learner in the proposed stacked ensemble approach to detect anomaly in credit card transactions. We apply stratified sampling technique and k-fold cross-validation process to address the issues of data imbalance and overfitting. In addition, the discordance rate is calculated to enhance the accuracy of ensemble learning performances. The proposed model is trained and tested using two datasets: CCF (Credit Card Fraud) and CCDP (Credit Card Default Payment). Experimental results demonstrate that our approach outperforms existing approaches, particularly in detecting anomalies from the minority class instances of these datasets.

Rapid runtime learning by curating small datasets of high-quality items obtained from memory.

We propose the "runtime learning" hypothesis which states that people quickly learn to perform unfamiliar tasks as the tasks arise by using task-relevant instances of concepts stored in memory during mental training. To make learning rapid, the hypothesis claims that only a few class instances are used, but these instances are especially valuable for training. The paper motivates the hypothesis by describing related ideas from the cognitive science and machine learning literatures. Using computer simulation, we show that deep neural networks (DNNs) can learn effectively from small, curated training sets, and that valuable training items tend to lie toward the centers of data item clusters in an abstract feature space. In a series of three behavioral experiments, we show that people can also learn effectively from small, curated training sets. Critically, we find that participant reaction times and fitted drift rates are best accounted for by the confidences of DNNs trained on small datasets of highly valuable items. We conclude that the runtime learning hypothesis is a novel conjecture about the relationship between learning and memory with the potential for explaining a wide variety of cognitive phenomena.

Iterative minority oversampling and its ensemble for ordinal imbalanced datasets

Ordinal classification of imbalanced datasets is a challenging problem that occurs in many real-world applications. The main challenge is to simultaneously consider the classes ordering and imbalanced distribution. Although the classic synthetic instances oversampling techniques can improve the identification of minority classes, they easily incur the damage of the classes ordering when the synthetic instances fall in non-adjacent classes regions. In this paper, we propose a powerful method for handling the imbalanced problem embedded in the ordinal classification, namely Iterative Minority oversampling technique for imbalanced Ordinal Classification (IMOC). Concretely, we first develop an iterative identification procedure to select the minority instance that is hardest to learn. Then, a weighted oversampling probability distribution that respects the ordinal nature is used to generate synthetic minority instances to balance the skewed distribution. Furthermore, two novel ensemble versions are developed to boost the capability of our proposed IMOC. In order to verify the effectiveness and robustness of our proposed methods, an extensive experimental study is carried out on a large number of datasets from real-world applications. The experimental results supported by proper statistical tests indicate that our proposed methods outperform state-of-the-art algorithms in terms of the most frequently used performance measures.

An exact algorithm for the pickup and delivery problem with crowdsourced bids and transshipment

This paper addresses a pickup and delivery problem with crowdsourced bids and transshipment (PDPCBT) in last-mile delivery, where all requests can be satisfied by either using the own vehicle fleet or outsourcing with a small compensation to crowdshippers through transshipment facilities. The crowdshippers show their willingness to deliver by submitting bids to the e-commerce company. To minimize both the travel cost of vehicles and the compensation of crowdshippers, the routes of vehicles and the selection of bids need to be optimized simultaneously. We formulate the PDPCBT into an arc-based formulation and a route-based formulation, where the latter is strengthened by the subset row inequalities. Based on the route-based formulation, we present a branch-and-price-and-cut algorithm to solve it exactly. To deal with two possible ways of serving requests, we first decompose the corresponding pricing problem into a shortest path problem and a knapsack problem, and then tackle them in the same bi-directional labeling algorithm framework. We also discuss acceleration techniques and implementation details to speed up the performance of the overall procedure. Computational experiments are conducted on a set of classic request instances, together with a set of randomly generated bid instances. With a time limit of two hours, numerical results validate the efficiency and effectiveness of the proposed algorithm. Finally, sensitivity analysis and managerial findings are also provided on the PDPCBT.