scholarly journals An Efficient Cost-Sensitive Feature Selection Using Chaos Genetic Algorithm for Class Imbalance Problem

2016 ◽  
Vol 2016 ◽  
pp. 1-9 ◽  
Author(s):  
Jing Bian ◽  
Xin-guang Peng ◽  
Ying Wang ◽  
Hai Zhang
Keyword(s):  
Genetic Algorithm ◽  
Feature Selection ◽  
Network Security ◽  
Classification Accuracy ◽  
Large Scale ◽  
Class Imbalance ◽  
Evaluation Function ◽  
Class Imbalance Problem ◽  
Imbalance Problem ◽  
Chaos Genetic Algorithm

In the era of big data, feature selection is an essential process in machine learning. Although the class imbalance problem has recently attracted a great deal of attention, little effort has been undertaken to develop feature selection techniques. In addition, most applications involving feature selection focus on classification accuracy but not cost, although costs are important. To cope with imbalance problems, we developed a cost-sensitive feature selection algorithm that adds the cost-based evaluation function of a filter feature selection using a chaos genetic algorithm, referred to as CSFSG. The evaluation function considers both feature-acquiring costs (test costs) and misclassification costs in the field of network security, thereby weakening the influence of many instances from the majority of classes in large-scale datasets. The CSFSG algorithm reduces the total cost of feature selection and trades off both factors. The behavior of the CSFSG algorithm is tested on a large-scale dataset of network security, using two kinds of classifiers: C4.5 andk-nearest neighbor (KNN). The results of the experimental research show that the approach is efficient and able to effectively improve classification accuracy and to decrease classification time. In addition, the results of our method are more promising than the results of other cost-sensitive feature selection algorithms.

scholarly journals A Cost-Sensitive Sparse Representation Based Classification for Class-Imbalance Problem

2016 ◽  
Vol 2016 ◽  
pp. 1-9
Author(s):  
Zhenbing Liu ◽  
Chunyang Gao ◽  
Huihua Yang ◽  
Qijia He
Keyword(s):  
Sparse Representation ◽  
Classification Accuracy ◽  
Class Imbalance ◽  
Misclassification Rate ◽  
Data Sets ◽  
Class Imbalance Problem ◽  
Misclassification Cost ◽  
Practical Applications ◽  
Imbalance Problem ◽  
Positive Class

Sparse representation has been successfully used in pattern recognition and machine learning. However, most existing sparse representation based classification (SRC) methods are to achieve the highest classification accuracy, assuming the same losses for different misclassifications. This assumption, however, may not hold in many practical applications as different types of misclassification could lead to different losses. In real-world application, much data sets are imbalanced of the class distribution. To address these problems, we propose a cost-sensitive sparse representation based classification (CSSRC) for class-imbalance problem method by using probabilistic modeling. Unlike traditional SRC methods, we predict the class label of test samples by minimizing the misclassification losses, which are obtained via computing the posterior probabilities. Experimental results on the UCI databases validate the efficacy of the proposed approach on average misclassification cost, positive class misclassification rate, and negative class misclassification rate. In addition, we sampled test samples and training samples with different imbalance ratio and use F-measure, G-mean, classification accuracy, and running time to evaluate the performance of the proposed method. The experiments show that our proposed method performs competitively compared to SRC, CSSVM, and CS4VM.

scholarly journals Combining feature selection and hybrid approach redefinition in handling class imbalance and overlapping for multi-class imbalanced

2021 ◽  
Vol 21 (3) ◽  
pp. 1513
Author(s):  
Hartono Hartono ◽  
Erianto Ongko ◽  
Yeni Risyani
Keyword(s):  
Feature Selection ◽  
Learning Algorithm ◽  
Hybrid Approach ◽  
Feature Selection Method ◽  
Class Imbalance ◽  
Poor Performance ◽  
Class Imbalance Problem ◽  
Imbalance Problem ◽  
Classifier Performance ◽  
F Measure

<span>In the classification process that contains class imbalance problems. In addition to the uneven distribution of instances which causes poor performance, overlapping problems also cause performance degradation. This paper proposes a method that combining feature selection and hybrid approach redefinition (HAR) method in handling class imbalance and overlapping for multi-class imbalanced. HAR was a hybrid ensembles method in handling class imbalance problem. The main contribution of this work is to produce a new method that can overcome the problem of class imbalance and overlapping in the multi-class imbalance problem.  This method must be able to give better results in terms of classifier performance and overlap degrees in multi-class problems. This is achieved by improving an ensemble learning algorithm and a preprocessing technique in HAR <span>using minimizing overlapping selection under SMOTE (MOSS). MOSS was known as a very popular feature selection method in handling overlapping. To validate the accuracy of the proposed method, this research use augmented R-Value, Mean AUC, Mean F-Measure, Mean G-Mean, and Mean Precision. The performance of the model is evaluated against the hybrid method (MBP+CGE) as a popular method in handling class imbalance and overlapping for multi-class imbalanced. It is found that the proposed method is superior when subjected to classifier performance as indicate with better Mean AUC, F-Measure, G-Mean, and precision.</span></span>

Sign in / Sign up

Export Citation Format

Share Document