A new hybrid filter/wrapper algorithm for feature selection in classification

Zhang, Jixiong; Xiong, Yanmei; Min, Shungeng

doi:10.1016/j.aca.2019.06.054

Cited by 75 publications

(37 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Faker and Dogdu [21] considered homogeneity metric as a measure to rank and remove the least ranked features. Zhang et al [22], proposed a hybrid filter and wrapper method where they created a subset of features with bootstrapping strategy. For each subset, classification accuracy is calculated to find the optimized subset.…”

Section: Hybrid Methodsmentioning

confidence: 99%

Mini-Batch Normalized Mutual Information: A Hybrid Feature Selection Method

Thejas

Joshi²,

Iyengar

et al. 2019

IEEE Access

View full text Add to dashboard Cite

Feature Selection has been a significant preprocessing procedure for classification in the area of Supervised Machine Learning. It is mostly applied when the attribute set is very large. The large set of attributes often tend to misguide the classifier. Extensive research has been performed to increase the efficacy of the predictor by finding the optimal set of features. The feature subset should be such that it enhances the classification accuracy by the removal of redundant features. We propose a new feature selection mechanism, an amalgamation of the filter and the wrapper techniques by taking into consideration the benefits of both the methods. Our hybrid model is based on a two phase process where we rank the features and then choose the best subset of features based on the ranking. We validated our model with various datasets, using multiple evaluation metrics. Furthermore, we have also compared and analyzed our results with previous works. The proposed model outperformed many existent algorithms and has given us good results. INDEX TERMS Feature selection, filter method, hybrid feature selection, normalized mutual information, mini batch K-means, random forest, wrapper method.

show abstract

Section: Hybrid Methodsmentioning

confidence: 99%

Mini-Batch Normalized Mutual Information: A Hybrid Feature Selection Method

Thejas

Joshi²,

Iyengar

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Feature selection is widely used to remove information redundancy and over-fitting in machine learning models [43]- [50]. The filtering method, wrapper method, and embedding method are three commonly used approaches to feature selection [51]- [53]. Filtering methods [52], [54] LGBM is one of the wrapper methods [57] to find the optimal features subset of the feature space via a specific classification model, which is associated with specific classification algorithms.…”

Section: Feature Selectionmentioning

confidence: 99%

Escherichia Coli DNA N-4-Methycytosine Site Prediction Accuracy Improved by Light Gradient Boosting Machine Feature Selection Technology

Wang

Ding

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Recently, several machine-learning-based DNA N-4-methycytosine (4mC) predictors have been developed to provide deeper insight into the biological functions and mechanisms of 4mC. However, the performance of the existing classifiers for identification of Escherichia coli DNA 4mC sites is inadequate. Here, we present a new support vector machine 4mC predictor, named iEC4mC-SVM, for Escherichia coli (E.coli) DNA 4mC site identification, optimized using light gradient boosting machine feature selection technology. The iEC4mC-SVM predictor had a 10-fold cross-validation accuracy of 85.4% and Jackknife crossvalidation accuracy of 84.9%. The 83.2% independent testing accuracy of iEC4mC-SVM was 1.0-6.5% higher than those of state-of-the-art E. coli DNA 4mC site predictors. A t-distributed stochastic neighbor embedding analysis confirmed that the prediction performance enhancement of iEC4mC-SVM was due to the light gradient boosting machine feature selection. INDEX TERMS Bioinformatics, DNA, machine learning, support vector machine, sequences.

show abstract

“…A study from [14] proposed a filter and wrapper based techniques which a new distance based evaluation function is utilized if the same class samples are attracted to each another, whereas different class samples are far apart, and a set of candidate feature subsets is determined using a weighted bootstrapping search strategy. In order to select the optimum features, specific classifier and cross validation were used to validate the performance.…”

Section: Related Research On Feature Selection Techniquesmentioning

confidence: 99%

“…Filter techniques is highly recommended due to its efficiency of computationally fastest in handling large datasets with the use of ranking and space searching technique. Furthermore, features that is independently weak but strong as a group can be identified, the redundant features can be eliminated and the highest correlated features with the output class can be determined by utilizing the hybridization of feature selection techniques [6,14,15].…”

Section: Related Research On Feature Selection Techniquesmentioning

confidence: 99%

Ensamble Based Multi Filters Algorithm for Tumor Classification in High Dimensional Microarray Dataset

Hamid¹

2019

IJATCSE

View full text Add to dashboard Cite

Prevalent adoption of machine learning has magnified its requirements in high dimensional microarray data classification. Due to explosive increase of data dimensionality, the existence of features redundancy and ambiguity directly leads to classification inaccuracy. Filter feature selection algorithms are capable to boost classification accuracy and diminish computational complexity by extracting relevant information through supervised learning. However, the independent filter algorithm is incompetent to consider the features interaction which resulting an imbalance selection of significant features and consequently degrading the classifier performance. This paper presents an assemblage of multi filters algorithm which assembles four filters algorithm outputs with frequency of occurrence rate evaluation to improve classification performance by attaining an optimal number of significant features. Experimental analysis was performed on a standard Breast Cancer dataset consists of 286 instances and Support Vector Machine (SVM) classifier. The experimental results proved that the ensemble based multi filters algorithm with occurrence rate evaluation successfully depletes from 9 original dataset features to 5 optimal significant features. The finding indicates that this technique competently signifies SVM classification performance in terms of accuracy with optimum significant features for high dimensional microarray data compared to independent filter algorithm.

show abstract

A new hybrid filter/wrapper algorithm for feature selection in classification

Cited by 75 publications

References 33 publications

Mini-Batch Normalized Mutual Information: A Hybrid Feature Selection Method

Mini-Batch Normalized Mutual Information: A Hybrid Feature Selection Method

Escherichia Coli DNA N-4-Methycytosine Site Prediction Accuracy Improved by Light Gradient Boosting Machine Feature Selection Technology

Ensamble Based Multi Filters Algorithm for Tumor Classification in High Dimensional Microarray Dataset

Contact Info

Product

Resources

About