Enhancing Robust and Stable Feature Selection Through the Integration of Ranking Methods and Wrapper Techniques in...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• High-dimensional data, common in microarray analysis, increases computational complexity and reduces classification performance.
• Accurate classification of genetic data is crucial for diagnosing diseases like cancer.
• Dimension reduction is essential for managing complex genetic datasets.

Purpose of the Study:

• To propose a robust method for dimension reduction and classification of genetic microarray data.
• To enhance the stability and accuracy of feature selection in high-dimensional biological datasets.
• To improve the performance of disease diagnosis models using genetic information.

Main Methods:

• A hybrid feature ranking method incorporating gene interactions was combined with a wrapper method for feature selection.
• A support vector machine (SVM) was utilized for the classification task.
• A solution for addressing class imbalance in training data was implemented prior to SVM classification.

Main Results:

• The proposed approach demonstrated robust feature selection across five microarray databases, with metrics ranging from 0.70 to 0.88.
• Classification accuracy achieved using the developed method ranged from 91% to 96%.
• The method effectively managed high-dimensional genetic data for improved diagnostic capabilities.

Conclusions:

• The integrated feature selection and classification method offers a significant improvement for high-dimensional genetic data analysis.
• The approach provides a stable and accurate solution for microarray data classification, particularly in disease diagnosis.
• Effective handling of class imbalance is critical for reliable classification of genetic datasets.