Article
Computer Science, Artificial Intelligence
Zhe Wang, Xu Chu, Dongdong Li, Hai Yang, Weichao Qu
Summary: This paper proposes a two-class cost-sensitive matrixized classification model CsMatMHKS based on information entropy to reduce the total misclassification cost. Experimental results show that CsMatMHKS not only reduces the sum of classification costs but also has comparable classification accuracy.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Information Systems
Toon Vanderschueren, Tim Verdonck, Bart Baesens, Wouter Verbeke
Summary: Predictive models are increasingly used to optimize decision-making and minimize costs. This work compared the predict-then-optimize approach with the predict-and-optimize approach in cost-sensitive classification. The key finding was that the decision-making strategy was generally more effective than training with a task-specific loss or their combination.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Interdisciplinary Applications
Yi Yang, Yuxuan Guo, Xiangyu Chang
Summary: The article introduces a novel angle-based cost-sensitive classification framework for multicategory classification without the sum-to-zero constraint, with loss functions proven to be Fisher consistent. Two cost-sensitive multicategory boosting algorithms derived from the framework show competitive classification performances in numerical experiments against other existing boosting approaches.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2021)
Article
Computer Science, Artificial Intelligence
George Petrides, Wouter Verbeke
Summary: The study introduces a unified framework for cost-sensitive ensemble methods, categorizing and comparing them, including extensions and generalizations for methods like AdaBoost, Bagging, and Random Forest.
DATA MINING AND KNOWLEDGE DISCOVERY
(2022)
Article
Environmental Sciences
Yijun Duan, Xin Liu, Adam Jatowt, Hai-tao Yu, Steven Lynden, Kyoung-Sook Kim, Akiyoshi Matono
Summary: Deep learning algorithms have gained popularity in remote sensing, particularly in the application of graph data. This paper presents a dual cost-sensitive graph convolutional network (DCSGCN) model to address the challenges of imbalanced long-tailed class distributions and classification bias. The DCSGCN uses cost as complementary information to correct the posterior probability and introduces a new method for computing node cost labels. Experimental results demonstrate the superior performance of DCSGCN compared to other baseline models on real-world graphs with imbalanced class distributions.
Article
Computer Science, Artificial Intelligence
Khalid Y. Aram, Sarah S. Lam, Mohammad T. Khasawneh
Summary: This research focuses on selecting relevant independent features for Support Vector Machine (SVM) classifiers in a cost-sensitive manner. The proposed linear cost-sensitive SVM embedded feature selection model demonstrated competitive performance.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Yingying Chen, Zijie Hong, Xiaowei Yang
Summary: This article introduces a cost-sensitive online adaptive kernel learning algorithm to address large-scale imbalanced classification problems. It proposes a misclassification cost to balance the accuracy between the minority class and the majority class. Experimental results demonstrate that the algorithm significantly improves classification performance on most large-scale imbalanced datasets.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Information Systems
Xiuhua Chen, Chen Gong, Jian Yang
Summary: This paper proposes a novel algorithm CSPU for PU learning, which addresses class imbalance by imposing different misclassification costs on different classes. The algorithm outperforms other comparators in dealing with minority classes.
INFORMATION SCIENCES
(2021)
Article
Engineering, Electrical & Electronic
Hao Li, Jianzhao Li, Yue Zhao, Maoguo Gong, Yujing Zhang, Tongfei Liu
Summary: This study proposed a cost-sensitive self-paced learning framework with adaptive regularization for the classification of image time series to address the issues of inaccurate labeled samples and transfer learning. Experimental results showed a significant improvement in classification accuracy achieved by the proposed algorithm.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Computer Science, Artificial Intelligence
Reem Alotaibi, Peter Flach
Summary: This paper investigates cost-sensitive classification methods for multi-label classification, adopting a simple but general thresholding method that is applicable to most classification algorithms. It explores the choice of single and multiple thresholds and proposes cost curves and scatter diagrams for performance evaluation. Experimental evaluation on 13 multi-label datasets demonstrates that adjusting a global threshold instead of per-label threshold does not lead to significant performance loss.
Article
Computer Science, Artificial Intelligence
Amine Bejaoui, Khalil Elkhalil, Abla Kammoun, Mohamed-Slim Alouini, Tareq Al-Naffouri
Summary: The article addresses the challenge of learning from imbalanced training data and proposes a multi-model selection approach to improve classifier performance through parameter optimization.
PATTERN RECOGNITION LETTERS
(2021)
Article
Computer Science, Information Systems
Xinmin Zhang, Saite Fan, Zhihuan Song
Summary: Fault classification is crucial in industrial process monitoring, but the imbalanced distribution of real-life datasets poses challenges. This paper proposes a novel reinforcement learning-based cost-sensitive classifier (RLCC) for imbalanced fault classification. RLCC utilizes a cost-sensitive learning strategy and a newly designed reward, trained through an alternating iterative approach.
SCIENCE CHINA-INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Zhiyong Li, Junfeng Zhang, Xiao Yao, Gang Kou
Summary: Credit scoring tools are often used by lenders to identify bad borrowers. A cost-sensitive ML-LightGBM algorithm is proposed in this paper to improve predictive accuracy, and it outperforms other predictive models in detecting fraudsters on online lending platforms.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Khalid Y. Aram, Sarah S. Lam, Mohammad T. Khasawneh
Summary: This article introduces the Alternated Sorting Method Genetic Algorithm (ASMGA), which is a hybrid wrapper-filter algorithm for simultaneous feature selection and model selection for Support Vector Machine (SVM) classifiers. ASMGA approximates a set of Pareto optimal feature subsets based on three objectives: cost-sensitive error rate, feature subset size, and Max-Margin Feature Selection (MMFS)-based estimates of feature relevance and redundancy. The proposed algorithm outperforms canonical GA and NSGA-II on benchmark datasets, showing the potential of ASMGA in cost-sensitive feature selection.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yi Yang, Shuai Huang, Wei Huang, Xiangyu Chang
Summary: This article introduces a method that combines cost-sensitive learning and privacy protection by incorporating weight constants and weight functions, proposing two privacy-preserving algorithms, and demonstrating that this general framework can reduce misclassification costs and meet privacy requirements. Theoretical research shows that the choice of weight constants and weight functions not only affects the algorithm's consistency properties, but also significantly interacts with privacy protection levels.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Review
Computer Science, Software Engineering
Orvila Sarker, Asangi Jayatilaka, Sherif Haggag, Chelsea Liu, M. Ali Babar
Summary: This study provides a comprehensive view of the challenges and critical success factors in the design, implementation, and evaluation stages of phishing education, training, and awareness (PETA). The findings highlight the need to address human-centric issues, bridge users' knowledge gaps, and adopt personalized approaches to enhance defense against phishing attacks.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Carlos Araujo, Meuse Oliveira Jr., Bruno Nogueira, Paulo Maciel, Eduardo Tavares
Summary: This paper proposes a method based on stochastic Petri nets for evaluating the consistency levels of storage systems based on NoSQL DBMS. The method takes into account different consistency levels and redundant nodes, and estimates the system's availability, throughput, and the probability of accessing the newest data. Experimental results demonstrate the practical feasibility of this approach.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Review
Computer Science, Software Engineering
L. Giamattei, A. Guerriero, R. Pietrantuono, S. Russo, I. Malavolta, T. Islam, M. Dinga, A. Koziolek, S. Singh, M. Armbruster, J. M. Gutierrez-Martinez, S. Caro-Alvaro, D. Rodriguez, S. Weber, J. Henss, E. Fernandez Vogelin, F. Simon Panojo
Summary: This article presents the results of a systematic study on the available monitoring tools for DevOps and microservices. It provides a classification and analysis of these tools, aiming to be a useful reference for researchers and practitioners in this field.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Jessica Diaz, Jorge Perez, Isaque Alves, Fabio Kon, Leonardo Leite, Paulo Meirelles, Carla Rocha
Summary: This paper presents empirical research on the structure of DevOps teams in software-producing organizations to better understand the organizational structure and characteristics of teams adopting DevOps. A theory of DevOps taxonomies is built through analysis, and its consistency with other taxonomies is tested.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Sinan Sigurd Tanilkan, Jo Erskine Hannay
Summary: When deciding to develop new software, it is important to have a clear understanding of the intended benefits. However, our research shows that stakeholders' understanding of benefits often fluctuates during the development process, leading to uncertainty. Therefore, we recommend focusing on helping practitioners embrace changes in their understanding of benefits.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Pingyan Wang, Shaoying Liu, Ai Liu, Wen Jiang
Summary: This paper presents an approach that combines static analysis tools and manual audits to effectively detect various types of security vulnerabilities. By using a special Petri net representation, the proposed method is able to assist in the detection of taint-style vulnerabilities.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Edgar Sarmiento-Calisaya, Julio Cesar Sampaio do Prado Leite
Summary: This research introduces an automated requirements analysis approach that combines natural language processing, Petri-nets, and visualization techniques to improve the quality of scenario-based specifications, identify defects, and anticipate inconsistencies.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Jian Hu
Summary: This paper proposes a two-stage trace matrix optimization method for fault localization, which addresses the challenges of coincidental correctness and data imbalance in the current trace matrix. Through extensive experiments, significant improvements in fault localization effectiveness are demonstrated.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Fan Zhang, Manman Peng, Yuanyuan Shen, Qiang Wu
Summary: This study proposes a novel method called HFEDR that utilizes the hierarchical features of Transformer models and reorganizes training data to improve code search performance. Experimental results demonstrate the effectiveness and rationality of the proposed approach.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Tong Wang, Bixin Li
Summary: Software architecture erosion has a negative impact on software quality, performance, and evolution cost. This paper proposes an approach called EsArCost to locate the causes of architecture erosion and estimate the repair cost of each erosion problem. Experimental results show that EsArCost can effectively and efficiently estimate repair costs.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Xiajing Wang, Rui Ma, Wei Huo, Zheng Zhang, Jinyuan He, Chaonan Zhang, Donghai Tian
Summary: This paper proposes a new potential-aware fuzzing scheme called SYNTONY that measures seed potential using multiple objectives and prioritizes promising seeds to increase the number of unique crashes and coverage. Experimental results show that SYNTONY outperforms other fuzzing tools and has high compatibility and expansibility.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Stefano Lambiase, Gemma Catolino, Fabiano Pecorelli, Damian A. Tamburri, Fabio Palomba, Willem-Jan van den Heuvel, Filomena Ferrucci
Summary: This paper contributes to the existing body of knowledge on factors affecting productivity in software development by studying the cultural and geographical dispersion of a development community. The results show that cultural and geographical dispersion significantly impact productivity, suggesting that managers and practitioners should consider these aspects throughout the software development lifecycle.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Elaine Venson, Bradford Clark, Barry Boehm
Summary: The software industry has been under pressure to adopt security practices and reduce software vulnerabilities. This study quantifies the effort required to develop secure software in increasing levels of rigor and scope and provides validated cost multipliers for practitioners to estimate proper resources for adopting security practices.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Yangyang Zhao, Mingyue Jiang, Yibiao Yang, Yuming Zhou, Hanjie Ma, Zuohua Ding
Summary: Previous studies have ignored the potential associations between modules involved in the same defect, and this comprehensive study explores the implications of intra-defect associations for defect prediction. The majority of defects occur across functions, with implicit dependencies between the modules. By considering intra-defect associations and merging modules, the proposed data processing approach significantly improves defect prediction performance.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)
Article
Computer Science, Software Engineering
Meira Levy, Irit Hadar
Summary: This research sheds new light on how students learn and practice hybrid work in educational settings through two educational studies. The findings show the benefits of new educational programs in fostering empathy and innovation among students, while also highlighting the challenges and opportunities in addressing real challenges.
JOURNAL OF SYSTEMS AND SOFTWARE
(2024)