Estimating Optimal Feature Subsets Using Mutual Information Feature Selector and Rough Sets

Foitong, Sombut; Rojanavasu, Pornthep; Attachoo, Boonwat; Pinngern, Ouen

doi:10.1007/978-3-642-01307-2_103

Cited by 5 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, rough sets are known for reducing the redundancy by preserving the discrimination power of the dataset. Few works using rough set concepts with information measures include (Zeng et al 2014;Maji and Pal 2009;Foitong et al 2009;Yang et al 2014;Qian and Liang 2008). Rough sets concept used with MI for feature selection is found to improve the cancer classification accuracy (Xu et al 2009).…”

Section: Ranking Metricsmentioning

confidence: 99%

A hybrid feature selection model based on improved squirrel search algorithm and rank aggregation using fuzzy techniques for biomedical data classification

Nagarajan

Babu

2021

Netw Model Anal Health Inform Bioinforma

View full text Add to dashboard Cite

Feature selection has gained its importance due to the voluminous nature of the data. Owing to the computational complexity of wrapper approaches, the poor performance of filtering techniques, and the classifier dependency of embedded approaches, hybrid approaches are more commonly used in feature selection. Hybrid approaches use filtering metrics to reduce the computational complexity of wrapper algorithms and are proved to yield better feature subset. Though filtering metrics select the features based on their significance, most of them are unstable and biased towards the metric used. Moreover, the choice of filtering metrics depends largely on the distribution of data and data types. Biomedical datasets contain features with different distribution and types adding to the complexity in the choice of filtering metric. We address this problem by proposing a stable filtering method based on rank aggregation in hybrid feature selection model with Improved Squirrel search algorithm for biomedical datasets. Our proposed model is compared with other well-known and state-of-the-art methods and the results prove that our model exhibited superior performance in terms of classification accuracy and computational time. The robustness of our proposed model is proved by conducting experiments on nine biomedical datasets and with three different classifiers.

show abstract

Section: Ranking Metricsmentioning

confidence: 99%

A hybrid feature selection model based on improved squirrel search algorithm and rank aggregation using fuzzy techniques for biomedical data classification

Nagarajan

Babu

2021

Netw Model Anal Health Inform Bioinforma

View full text Add to dashboard Cite

show abstract

“…The induced evaluation function has been extensively applied to fuzzy rough sets based feature selection 37,48 , variable precision rough sets based feature selection 8 , dominance rough sets based feature selection 16 , rough sets and Bayesian networks based feature selection 40 , dynamic mutual information based feature selection 23 , and normalized mutual information based feature selection 7 . However, mutual information of knowledge relies strongly on prior probability of knowledge that is unknown in information systems.…”

Section: Introductionmentioning

confidence: 99%

Feature Selection in Decision Systems Based on Conditional Knowledge Granularity

Deng¹,

Yang²,

Hu³

2011

International Journal of Computational Intelligence Systems

View full text Add to dashboard Cite

Feature selection is an important technique for dimension reduction in machine learning and pattern recognition communities. Feature evaluation functions play essential roles in constructing feature selection algorithms. This paper introduces a new notion of knowledge granularity, called conditional knowledge granularity, reflecting relationship between conditional attributes and decision attribute. An evaluation function to measure significance of conditional attributes is proposed and equivalent characterization of attribute reduction is established based on the conditional knowledge granularity. An optimal algorithm for feature selection is developed on the basis of the proposed evaluation function. Furthermore, a novel approach to performing feature selection in an inconsistent decision system is put forward through establishing a rough communication between the inconsistent decision system and a consistent decision system. Simulated experiments verifies feasibility and efficiency of the proposed technique.

show abstract

Self-taught dimensionality reduction on the high-dimensional small-sized data

Zhu

Huang

Yang

et al. 2013

Pattern Recognition

131

View full text Add to dashboard Cite

Estimating Optimal Feature Subsets Using Mutual Information Feature Selector and Rough Sets

Cited by 5 publications

References 7 publications

A hybrid feature selection model based on improved squirrel search algorithm and rank aggregation using fuzzy techniques for biomedical data classification

A hybrid feature selection model based on improved squirrel search algorithm and rank aggregation using fuzzy techniques for biomedical data classification

Feature Selection in Decision Systems Based on Conditional Knowledge Granularity

Self-taught dimensionality reduction on the high-dimensional small-sized data

Contact Info

Product

Resources

About