Article
Computer Science, Artificial Intelligence
Jianming Liu, Jie Lei, Zhouyu Liao, Jiali He
Summary: This study proposes a novel software defect prediction model based on a twin support vector machine to address the issue of imbalanced data classification, achieving higher accuracy and robustness in classifying imbalanced data.
Article
Computer Science, Hardware & Architecture
Zhanyu Yang, Lu Lu, Quanyi Zou
Summary: Rank-oriented software defect prediction (ROSDP) aims to establish a model for predicting the testing priority of software modules based on defect severity, to allocate testing resources effectively. Some methods use linear models to predict the priority, which may have limited ranking performance in software repositories. To overcome this limitation, ensemble kernel-mapping-based ranking support vector machine (EKMRSVM) is developed, which uses ranking SVM to build a nonlinear ranking function approximated by the kernel-mapping method. Experimental results on 20 open source datasets show that EKMRSVM with kernel mapping is effective in improving performance and ensemble learning reduces time costs while ensuring ranking performance. Compared to baseline methods, EKMRSVM with an appropriate kernel function achieves better ranking performance.
IEEE TRANSACTIONS ON RELIABILITY
(2023)
Article
Multidisciplinary Sciences
Vinod Kumar Kulamala, Lov Kumar, Durga Prasad Mohapatra
Summary: Software fault prediction is a process that helps to identify fault-prone modules in early stages of software development, with the goal of improving software quality with optimized effort and cost. This study explores Least Square Support Vector Machines (LSSVM) and compares software fault prediction models using different kernels. Experimental results show that LSSVM with polynomial kernel performs better than LSSVM with linear kernel and is similar to RBF kernel.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2021)
Article
Computer Science, Information Systems
Suneel Kumar Rath, Madhusmita Sahu, Shom Prasad Das, Sukant Kishoro Bisoy, Mangal Sain
Summary: Software defect prediction aims to predict potential flaws in new software modules in advance by creating an effective prediction model. However, unnecessary and duplicated features can affect the model's performance. This research proposes an SVM and ELM-based algorithm for software reliability prediction and investigates factors that influence prediction accuracy. Experimental results demonstrate that the ELM-based model achieves higher prediction accuracy.
Article
Computer Science, Artificial Intelligence
Kechao Wang, Lin Liu, Chengjun Yuan, Zhifei Wang
Summary: This paper investigates a software defect prediction model based on LASSO-SVM, combining feature selection and support vector machine to improve prediction accuracy. Simulation results demonstrate that the proposed model outperforms traditional defect prediction models, with faster prediction speed.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Geosciences, Multidisciplinary
Hoang Nguyen, Xuan-Nam Bui, Yosoon Choi, Chang Woo Lee, Danial Jahed Armaghani
Summary: This study proposed a novel data-driven model for estimating fly-rock distance in bench blasting in open-pit mines using a combination of the whale optimization algorithm (WOA), support vector machine (SVM) and kernel functions. The WOA-SVM-RBF model showed the highest accuracy in predicting the fly-rock distance among all models investigated.
NATURAL RESOURCES RESEARCH
(2021)
Article
Computer Science, Theory & Methods
Yunhua Zhao, Kostadin Damevski, Hui Chen
Summary: In recent years, there has been a sustained focus on research on software defect prediction, with a specific focus on Just-in-Time Software Defect Prediction (JIT-SDP) that aims to predict the likelihood of each incremental software change being defective. A systematic survey of 67 JIT-SDP studies has been conducted to summarize best practices, carry out a meta-analysis, and suggest future research directions. The meta-analysis indicates that JIT-SDP's predictive performance is most effective in projects with relatively high defect ratios. Future research directions include situating techniques in their application domain, reliability-aware JIT-SDP, and user-centered JIT-SDP.
ACM COMPUTING SURVEYS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Mohammad Azzeh, Ali Bou Nassif, Shadi Banitaan
Summary: This paper investigates the performance of Support Vector Machine (SVM) in software defect prediction and finds that the choice of kernel functions and feature dimensionality reduction can affect its accuracy and stability. The experiments reveal that no kernel function can maintain stability across different settings, while Radial basis and Sigmoid kernel functions exhibit relatively more stability.
2022 21ST IEEE INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS, ICMLA
(2022)
Article
Environmental Sciences
Chih-Chun Liu, Tzu-Chi Lin, Kuang-Yu Yuan, Pei-Te Chiueh
Summary: This study utilizes the support vector machine (SVM) method to predict air quality and considers geographic features and time series. The results show high accuracy for short-term temporal prediction, with meteorological and climatic factors influencing seasonal differences. In the spatial inference stage, urbanization and city types were found to impact air quality, while agriculture and forest use, transportation use, residential use, and economic factors were correlated with AQIs.
Article
Optics
Hongwei Li, Hailiang Chen, Yuxin Li, Qiang Chen, Xiaoya Fan, Shuguang Li, Mingjian Ma
Summary: In this paper, support vector machines (SVMs) based on radial basis functions were used to predict the optical properties of photonic crystal fiber (PCF). Well-trained SVMs can accurately and quickly predict the effective refractive index, chromatic dispersion, and confinement loss of PCF. Compared to artificial neural networks (ANNs), SVMs are more accurate and show stable prediction results.
Article
Computer Science, Information Systems
Jinseong Park, Yujin Choi, Junyoung Byun, Jaewook Lee, Saerom Park
Summary: In this paper, a multi-class classification method using kernel supports and a dynamical system under differential privacy is proposed. For small datasets, kernel methods, such as kernel support vector machines (SVMs), show good generalization performance with high-dimensional feature mapping. However, kernel SVMs have a fundamental weakness in achieving differential privacy because they construct decision functions based on a subset of the training data called support vectors. To address these limitations, a two-phase classification algorithm based on support vector data description (SVDD) is developed. It generates a differentially private SVDD (DP-SVDD) by perturbing the sphere center in a high-dimensional feature space and partitions the input space using a dynamical system for classification.
INFORMATION SCIENCES
(2023)
Article
Mathematics
Maria Jose Hernandez-Molinos, Angel J. Sanchez-Garcia, Rocio Erandi Barrientos-Martinez, Juan Carlos Perez-Arriaga, Jorge Octavio Ocharan-Hernandez
Summary: Software defect prediction is an important area in software engineering, aiming to identify and fix problems before they become costly and hard-to-fix bugs. This research evaluates three algorithms to build Bayesian Networks for classifying project defects and compares the results with other commonly used approaches. The results show that Bayesian algorithms exhibit less variability and provide greater robustness in software defect predictions compared to Decision Tree and Random Forest.
Article
Computer Science, Artificial Intelligence
Gherardo Varando, Salvador Catsis, Emiliano Diaz, Gustau Camps-Valls
Summary: Bivariate causal discovery is the task of inferring the causal relationship between two random variables from observational data. This paper proposes an ensemble algorithm that combines classical and data-driven methods, achieving superior performance on various synthetic and real-world problems.
APPLIED SOFT COMPUTING
(2024)
Article
Computer Science, Artificial Intelligence
Yinan Guo, Zirui Zhang, Fengzhen Tang
Summary: Feature selection is important in machine learning to reduce complexity and simplify interpretation. A novel non-linear method proposed in this paper uses kernelized multi-class support vector machines and fast recursive feature elimination to select features that work well for all classes, resulting in lower computational time complexity.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Artificial Intelligence
Hossein Moosaei, M. A. Ganaie, Milan Hladik, M. Tanveer
Summary: Imbalanced datasets are common in real-world problems. Traditional classification algorithms have limitations in handling imbalanced data. To improve classification performance on imbalanced datasets, an improved reduced universum twin support vector machine (IRUTSVM) algorithm is proposed, which introduces new constraints and reduces computational time.
Article
Computer Science, Software Engineering
Cuauhtemoc Lopez-Martin, Yenny Villuendas-Rey, Mohammad Azzeh, Ali Bou Nassif, Shadi Banitaan
JOURNAL OF SYSTEMS AND SOFTWARE
(2020)
Review
Computer Science, Software Engineering
Mohammad Azzeh, Ali Bou Nassif, Imtinan Basem Attili
Summary: The use of Use Case Points (UCP) method for predicting software project effort is gaining popularity. However, the area has not been systematically reviewed, highlighting the need for a literature review to guide and support effort estimation research. The current study aims to classify and analyze UCP effort estimation papers based on various criteria and perspectives, to explore accuracy, estimation context, and the impact of combined techniques on UCP accuracy.
SCIENCE OF COMPUTER PROGRAMMING
(2021)
Article
Computer Science, Software Engineering
Mohammad Azzeh, Ali Bou Nassif, Cuauhtemoc Lopez-Martin
Summary: This paper investigates the impact of data locality approaches on productivity and effort prediction based on multiple UCP variables. It also explores the relationship between productivity and other UCP variables.
SOFTWARE QUALITY JOURNAL
(2021)
Article
Computer Science, Software Engineering
Yousef Alqasrawi, Mohammad Azzeh, Yousef Elsheikh
Summary: Estimating software effort has been a challenge, and this paper introduces a more sophisticated locality approach called Locally Weighted Regression (LWR) to learn from local data and build estimation models with multiple local regression models.
SCIENCE OF COMPUTER PROGRAMMING
(2022)
Article
Computer Science, Software Engineering
Mohammad Azzeh, Ali Bou Nassif, Yousef Elsheikh, Lefteris Angelis
Summary: Estimating software effort using Use Case Points involves considering productivity, with challenges in predicting it and the aid of historical data. Learning productivity from historical data is more effective than traditional methods, and environmental factors may not accurately predict productivity.
SCIENCE OF COMPUTER PROGRAMMING
(2022)
Article
Computer Science, Information Systems
Yousef Elsheikh, Yousef Alqasrawi, Mohammad Azzeh
Summary: This study addresses the issue of selecting and prioritizing strategies for the successful implementation of e-government programs. By using different ranking methods and measurement criteria, the study highlights the importance of obtaining stable rankings in selecting the best strategies.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Shaima Hameed, Yousef Elsheikh, Mohammad Azzeh
Summary: Software development companies have faced long-standing challenges in accurately estimating the effort required for software projects. However, research has shown that machine learning techniques, such as case-based reasoning, can improve accuracy. The case-based reasoning technique, though effective, has difficulty in tuning its multiple parameters. This paper proposes the use of a genetic algorithm to find the best combination of parameters and improve accuracy. The results show the effectiveness of this approach, which is beneficial for project managers in financial planning and cost control.
INFORMATION AND SOFTWARE TECHNOLOGY
(2023)
Article
Biology
Fatma Hilal Yagin, Ipek Balikci Cicek, Abedalrhman Alkhateeb, Burak Yagin, Cemil Colak, Mohammad Azzeh, Sami Akbulut
Summary: This study presents a model that applies explainable artificial intelligence (XAI) methods to assist in diagnosing COVID-19. By using machine learning techniques on COVID-19 metagenomic next-generation sequencing (mNGS) samples and combining LIME and SHAP for explanations, the model successfully predicts COVID-19 and identifies biomarker candidate genes associated with the disease.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Computer Science, Software Engineering
Mohammad Azzeh, Yousef Alqasrawi, Yousef Elsheikh
Summary: Defect density is crucial for software testing and maintenance, used to distribute limited human resources effectively. We propose a new prediction model that integrates gray system theory and fuzzy logic to handle uncertainty in software measurement. The model's performance was validated against public defect datasets, outperforming other prediction models, especially with high sparsity ratios.
JOURNAL OF SOFTWARE-EVOLUTION AND PROCESS
(2023)
Article
Multidisciplinary Sciences
Omar Abboushi, Mohammad Azzeh
Summary: This study fine-tuned AraGPT2, the most advanced Arabic pre-trained transformer, on a large poetry corpus to generate Arabic poems with specific meter and rhyme. The results showed high-quality generated poems according to standard evaluation and expert evaluation. However, concerns were raised regarding potential misuse of this technology.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2023)
Article
Operations Research & Management Science
Sinan Faouri, Mahmood AlBashayreh, Mohammad Azzeh
Summary: This study investigates the stability of machine learning algorithms for dementia prediction. Through numerous experiments, it is found that support vector machine and Naive Bayes are the most stable algorithms, and using Information Gain (IG) appears to be more effective than using Principal Component Analysis (PCA) for predicting dementia.
DECISION SCIENCE LETTERS
(2022)
Article
Computer Science, Information Systems
Firas Alghanim, Mohammad Azzeh, Ammar El-Hassan, Hazem Qattous
Summary: Delivering a reliable and high-quality software system to clients is a challenging task, and defect density is a key measure of system quality. However, predicting defect density before testing the modules is time-consuming. To address this issue, managers can build prediction models using deep learning to detect defective modules, thus reducing testing costs and improving resource utilization. Our study demonstrates that deep learning is effective in handling sparse data and outperforms other machine learning methods in datasets with high and very high sparsity ratios, while also being a competitive choice for datasets with medium or low sparsity ratios.
Article
Automation & Control Systems
Mohammad AL-Oudat, Saleh Alomari, Hazem Qattous, Mohammad Azzeh, Tariq AL-Munaize
Summary: The study introduces a vision-based model for initial diagnosis of the biliary tree, utilizing different image processing methods to segment MRI images and extract features to determine patients' health conditions. The research, using a database of 200 MRI images, demonstrates the effectiveness of extracted features with various classifiers.
INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL
(2021)
Article
Automation & Control Systems
Ahmad Alqwadri, Mohammad Azzeh, Fadi Almasalha
Summary: This paper proposes a new reputation system using machine learning to predict consumer reliability and compute product reputation score. The model is evaluated on MovieLens benchmarking datasets and compared to previous rating aggregation models, showing promising results and potential as a solution for reputation systems. The proposed approach can be integrated with online recommendation systems to enhance user experience in online shopping markets.
INTERNATIONAL JOURNAL OF AUTOMATION AND COMPUTING
(2021)
Article
Computer Science, Software Engineering
Amel Mammar, Meriem Belguidoum, Saddam Hocine Hiba
Summary: This paper introduces a formal EVENT-B-based approach for modeling and verifying the deployment of component-based applications. By gradually refining an abstract model, a precise specification is built, and mathematical reasoning is used to prove its correctness. The presented approach validates the deployment in a cloud environment using PROB and ensures the construction of a correct system that meets the constraints.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Shuqi Liu, Yu Zhou, Longbing Ji, Tingting Han, Taolue Chen
Summary: In this paper, we propose a framework that combines GUI events deduplication with an adaptive semantic matching strategy to enhance the usability of reused tests. Experimental evaluation demonstrates that the framework improves widget mapping performance, significantly reduces event redundancy, and reduces the manual effort of creating tests for similar applications.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Xiangyu Mu, Lei Liu, Peng Zhang, Jingyao Li, Hui Li
Summary: The aim of this study is to reduce the size of the test case set required to detect the commutativity problem of the reduce function. By determining the pattern of the function and selecting corresponding test cases, the proposed test case generation strategy can achieve the same accuracy with a smaller test case set. It has been shown to be effective and has a high recall rate.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Padmalata Nistala, Asha Rajbhoj, Vinay Kulkarni, Sapphire Noronha, Ankit Joshi
Summary: This paper presents an automated proposal development approach using a combination of model-based and AI-enabled techniques, and discusses the successful deployment and user feedback of the system.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Jacco O. G. Krijnen, Manuel M. T. Chakravarty, Gabriele Keller, Wouter Swierstra
Summary: Compiler correctness is a long-standing problem, and it becomes more significant with the rise of smart contracts on blockchains. A translation certification framework can address the trust issue for low-level code on the blockchain, allowing users to have confidence in the compilation process of smart contracts.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Phillip James, Faron Moller, Filippos Pantekis
Summary: OnTrack is a tool that supports railway verification workflows using model driven engineering frameworks, allowing railway engineers to interact with verification procedures through encapsulating formal methods.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Oleg Kiselyov
Summary: Heterogeneous metaprogramming systems leverage higher-level host languages to generate lower-level object language code, enabling faster production of high-performant code with correctness guarantees. This paper presents two systems with OCaml as the host language and C as the object language, discussing their implementation and applications.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Conor Reynolds, Rosemary Monahan
Summary: This paper provides a detailed approach to formalize a fragment of the theory of institutions in the Coq proof assistant. The approach is illustrated and evaluated by instantiating the framework with specific institution examples.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Herbert Rausch Fernandes, Giovanni Freitas Gomes, Antonio Carlos Pinheiro de Oliveira, Sergio Vale Aguiar Campos
Summary: Alzheimer's disease is a common form of dementia with no effective drug treatment available. In this study, a statistical model checking approach was used to analyze protein and drug interactions and evaluate the effects of different drugs on the components contributing to Alzheimer's disease. The results showed that rapamycin could slow down the biological process causing neuronal death, while LY294002 and NVP-BEZ235 may increase tau phosphorylation. These findings provide important insights for the scientific community and raise awareness about potential side effects of PI3K inhibitor drugs.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Erwan Mahe, Christophe Gaston, Pascale Le Gall
Summary: This paper presents an Interaction Language to encode Sequence Diagrams (SD) and associates it with three different formal semantics. This allows for direct formal verification of SD, while preserving traceability of SD concepts and executed actions, and addressing the translation of problematic operators.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Joan Giner-Miguelez, Abel Gomez, Jordi Cabot
Summary: Datasets are crucial for training and evaluating machine learning models, but they can also lead to undesirable behaviors like biased predictions. To tackle this issue, the machine learning community suggests adopting consistent guidelines for dataset descriptions. However, these guidelines rely on natural language descriptions, which hinder automated computation and analysis. To overcome this, we present DescribeML, a language engineering tool that provides precise, structured descriptions of machine learning datasets, including their composition, provenance, and social concerns.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Andrey Sadovykh, Bilal Said, Dragos Truscan, Hugo Bruneliere
Summary: In this paper, the authors report on their 7 years of practical experience with an iterative Model-based Requirements Engineering (MBRE) approach and language in five large European collaborative projects. They demonstrate through significant data sets that this model-based approach provides interesting benefits in terms of scalability, heterogeneity, adaptability, traceability, automation, consistency and quality, and usefulness or usability. Concrete examples from these projects are provided to illustrate the application of the MBRE approach and language, and the authors discuss the general benefits and limitations of using such an approach, as well as the lessons learned over the years.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Alfa Yohannis, Dimitris Kolovos, Antonio Garcia-Dominguez
Summary: Picto Web is a multi-tenant web-based tool that allows exploration of complex models by transforming them into various transient web-based views using rule-based transformations. It uses a lazy view computation approach to efficiently support large models and complex transformations, and includes monitoring and push notification facilities for automatic recomputation of views and updated delivery to clients.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Enes Yigitbas, Maximilian Schmidt, Antonio Bucchiarone, Sebastian Gottschalk, Gregor Engels
Summary: UML has become a popular modeling language used in computer science courses, and various interactive learning applications have been developed to improve student engagement and learning outcomes. However, these applications have not successfully created immersive environments for students. Therefore, this study introduces GaMoVR, a VR-based and gamified learning environment, which provides an interactive and fun learning experience for students learning about UML modeling.
SCIENCE OF COMPUTER PROGRAMMING
(2024)
Article
Computer Science, Software Engineering
Yaxin Zhao, Lina Gong, Wenhua Yang, Yu Zhou
Summary: Accessible design aims to enable as many people as possible to access software products and services. This study investigates the interaction between accessibility issues and other factors affecting software performance. By analyzing a large number of accessibility issues, the study reveals the characteristics of these issues and their relationship with software quality attributes.
SCIENCE OF COMPUTER PROGRAMMING
(2024)