Related Experiment Video
Updated: Feb 23, 2026
Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
A filter feature selection method based on the Maximal Information Coefficient and Gram-Schmidt Orthogonalization for
Hongqiang Lyu1, Mingxi Wan2, Jiuqiang Han3
1School of Electronic and Information Engineering, Xi'an Jiaotong University, Xi'an, 710049, PR China; School of Life Science and Technology, Xi'an Jiaotong University, Xi'an, 710049, PR China.
Abstract:
A filter feature selection technique has been widely used to mine biomedical data. Recently, in the classical filter method minimal-Redundancy-Maximal-Relevance (mRMR), a risk has been revealed that a specific part of the redundancy, called irrelevant redundancy, may be involved in the minimal-redundancy component of this method. Thus, a few attempts to eliminate the irrelevant redundancy by attaching additional procedures to mRMR, such as Kernel Canonical Correlation Analysis based mRMR (KCCAmRMR), have been made. In the present study, a novel filter feature selection method based on the Maximal Information Coefficient (MIC) and Gram-Schmidt Orthogonalization (GSO), named Orthogonal MIC Feature Selection (OMICFS), was proposed to solve this problem. Different from other improved approaches under the max-relevance and min-redundancy criterion, in the proposed method, the MIC is used to quantify the degree of relevance between feature variables and target variable, the GSO is devoted to calculating the orthogonalized variable of a candidate feature with respect to previously selected features, and the max-relevance and min-redundancy can be indirectly optimized by maximizing the MIC relevance between the GSO orthogonalized variable and target. This orthogonalization strategy allows OMICFS to exclude the irrelevant redundancy without any additional procedures. To verify the performance, OMICFS was compared with other filter feature selection methods in terms of both classification accuracy and computational efficiency by conducting classification experiments on two types of biomedical datasets. The results showed that OMICFS outperforms the other methods in most cases. In addition, differences between these methods were analyzed, and the application of OMICFS in the mining of high-dimensional biomedical data was discussed. The Matlab code for the proposed method is available at https://github.com/lhqxinghun/bioinformatics/tree/master/OMICFS/.
Insights
A new Orthogonal MIC Feature Selection (OMICFS) method effectively removes irrelevant redundancy in biomedical data mining. OMICFS improves feature selection accuracy and computational efficiency compared to existing methods.
Area of Science:
- Biomedical Data Mining
- Bioinformatics
- Machine Learning
Background:
- Filter feature selection is crucial for biomedical data mining.
- The minimal-Redundancy-Maximal-Relevance (mRMR) method faces challenges with irrelevant redundancy.
- Existing mRMR enhancements, like KCCAmRMR, require additional procedures.
Purpose of the Study:
- To propose a novel filter feature selection method, OMICFS, addressing irrelevant redundancy.
- To enhance feature selection by optimizing max-relevance and min-redundancy criteria.
- To improve classification accuracy and computational efficiency in biomedical data analysis.
Main Methods:
- Utilizing Maximal Information Coefficient (MIC) for feature-target relevance.
- Employing Gram-Schmidt Orthogonalization (GSO) to create orthogonalized features.
- Indirectly optimizing relevance and redundancy by maximizing MIC with GSO-orthogonalized variables.
Main Results:
- OMICFS effectively excludes irrelevant redundancy without extra steps.
- Demonstrated superior performance in classification accuracy across biomedical datasets.
- Showcased improved computational efficiency compared to other filter methods.
Conclusions:
- OMICFS offers a robust solution for irrelevant redundancy in feature selection.
- The method is suitable for mining high-dimensional biomedical data.
- OMICFS presents a significant advancement in filter-based feature selection techniques.
