An empirical study on the joint impact of feature selection and data resampling on imbalance classification

Zhang, Chongsheng; Soda, Paolo; Bi, Jingjun; Fan, Gaojuan; Almpanidis, George; García, Salvador

doi:10.1007/s10489-022-03772-1

Cited by 19 publications

(11 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [16], researchers proposed a framework that investigated the joint effects of FS and data resampling on unbalanced two-class classification. The research compared two contrasting approaches, i.e., one in which FS was applied before resampling the data (FS+DS) and another in which FS was applied after resampling the data (DS+FS).…”

Section: ) Embedded Methodmentioning

confidence: 99%

“…In [15], seven classifiers were assessed, specifically KNN, NB, Linear Discriminant Analysis (LDA), Linear Regression, DT, and SVM as individual classifiers while RF was employed as an ensemble classifier. In [16], three distinct classifiers from diverse families were utilized; including C4.5, SVM, and MLP aiming to evaluate the effectiveness of the proposed framework. A distribution of primary studies over individual classifiers is shown in Figure 9 and distribution of primary studies over ensemble classifiers is presented in Figure 10.…”

Section: Embedded Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Analysis of Feature Selection Methods in Software Defect Prediction Models

Ali,

Mazhar,

Shahzad

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Improving software quality by proactively detecting potential defects during development is a major goal of software engineering. Software defect prediction plays a central role in achieving this goal. The power of data analytics and machine learning allows us to focus our efforts where they are needed most. A key factor in the success of software fault prediction is selecting relevant features and reducing data dimensionality. Feature selection methods contribute by filtering out the most critical attributes from a plethora of potential features. These methods have the potential to significantly improve the accuracy and efficiency of fault prediction models. However, the field of feature selection in the context of software fault prediction is vast and constantly evolving, with a variety of techniques and tools available. Based on these considerations, our systematic literature review conducts a comprehensive investigation of feature selection methods used in the context of software fault prediction. The research uses a refined search strategy involving four reputable digital libraries, including IEEE Explore, Science Direct, ACM Digital Library, and Springer Link, to provide a comprehensive and exhaustive review through a rigorous analysis of 49 selected primary studies from 2014. The results highlight several important issues. First, there is a prevalence of filtering and hybrid feature selection methods. Second, single classifiers such as Naïve Bayes, Support Vector Machine, and Decision Tree, as well as ensemble classifiers such as Random Forest, Bagging, and AdaBoost are commonly used. Third, evaluation metrics such as area under the curve, accuracy, and F-measure are commonly used for performance evaluation. Finally, there is a clear preference for tools such as WEKA, MATLAB, and Python. By providing insights into current trends and practices in the field, this study offers valuable guidance to researchers and practitioners to make informed decisions to improve software fault prediction models and contribute to the overall improvement of software quality.

show abstract

Section: ) Embedded Methodmentioning

confidence: 99%

Section: Embedded Methodsmentioning

confidence: 99%

Analysis of Feature Selection Methods in Software Defect Prediction Models

Ali,

Mazhar,

Shahzad

et al. 2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Although they are computationally expensive, wrapper methods like Recursive Feature Elimination often outperform filter methods. The downside is that they are prone to overfitting and are not suitable for high-dimensional datasets due to computational intensity levels 8 – 10 .…”

Section: In-depth Review Of Existing Machine Learning Models Used For...mentioning

confidence: 99%

Feature importance feedback with Deep Q process in ensemble-based metaheuristic feature selection algorithms

Potharlanka,

2024

Sci Rep

View full text Add to dashboard Cite

Feature selection is an indispensable aspect of modern machine learning, especially for high-dimensional datasets where overfitting and computational inefficiencies are common concerns. Traditional methods often employ either filter, wrapper, or embedded approaches, which have limitations in terms of robustness, computational load, or capability to capture complex interactions among features. Despite the utility of metaheuristic algorithms like Particle Swarm Optimization (PSO), Firefly Algorithm (FA), and Whale Optimization (WOA) in feature selection, there still exists a gap in efficiently incorporating feature importance feedback into these processes. This paper presents a novel approach that integrates the strengths of PSO, FA, and WOA algorithms into an ensemble model and further enhances its performance by incorporating a Deep Q-Learning framework for relevance feedbacks. The Deep Q-Learning module intelligently updates feature importance based on model performance, thereby fine-tuning the selection process iteratively. Our ensemble model demonstrates substantial gains in effectiveness over traditional and individual metaheuristic approaches. Specifically, the proposed model achieved a 9.5% higher precision, an 8.5% higher accuracy, an 8.3% higher recall, a 4.9% higher AUC, and a 5.9% higher specificity across multiple software bug prediction datasets and samples. By resolving some of the key issues in existing feature selection methods and achieving superior performance metrics, this work paves the way for more robust and efficient machine learning models in various applications, from healthcare to natural language processing scenarios. This research provides an innovative framework for feature selection that promises not only superior performance but also offers a flexible architecture that can be adapted for a variety of machine learning challenges.

show abstract

“…On account of the above debates, Data Mining (DM) conceptions and principles could be exploited to locate the perceptive patterns behind the raw datasets [ 5 ]. DM is a general process that consists of a series of transformation steps, from a pre-processing step of the data to a post-processing step of the results generated by a pre-built DM method [ 6 ]. Thus, data preparation methods are a basic request to fetch the most out of the data, and thus create a classification model with an optimum search performance.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Ali Baba and the forty thieves algorithm for feature selection

Braik

2022

Neural Comput & Applic

View full text Add to dashboard Cite

Feature Selection (FS) aims to ameliorate the classification rate of dataset models by selecting only a small set of appropriate features from the initial range of features. In consequence, a reliable optimization method is needed to deal with the matters involved in this problem. Often, traditional methods fail to optimally reduce the high dimensionality of the feature space of complex datasets, which lead to the elicitation of weak classification models. Meta-heuristics can offer a favorable classification rate for high-dimensional datasets. Here, a binary version of a new human-based algorithm named Ali Baba and the Forty Thieves (AFT) was applied to tackle a pool of FS problems. Although AFT is an efficient meta-heuristic for optimizing many problems, it sometimes exhibits premature convergence and low search performance. These issues were mitigated by proposing three enhanced versions of AFT, namely: (1) A Binary Multi-layered AFT called BMAFT which uses hierarchical and distributed frameworks, (2) Binary Elitist AFT (BEAFT) which uses an elitist learning strategy, and, (3) Binary Self-adaptive AFT (BSAFT) which uses an adapted tracking distance parameter. These versions along with the basic Binary AFT (BAFT) were expansively assessed on twenty-four problems gathered from different repositories. The results showed that the proposed algorithms substantially enhance the performance of BAFT in terms of convergence speed and solution accuracy. On top of that, the overall results showed that BMAFT is the most competitive, which provided the best results with excellent performance scores compared to other competing algorithms.

show abstract

An empirical study on the joint impact of feature selection and data resampling on imbalance classification

Cited by 19 publications

References 28 publications

Analysis of Feature Selection Methods in Software Defect Prediction Models

Analysis of Feature Selection Methods in Software Defect Prediction Models

Feature importance feedback with Deep Q process in ensemble-based metaheuristic feature selection algorithms

Enhanced Ali Baba and the forty thieves algorithm for feature selection

Contact Info

Product

Resources

About