Chaotic genetic algorithm for gene selection and classification problems.

Area of Science:

• Bioinformatics
• Computational Biology
• Machine Learning

Background:

• High-dimensional microarray data presents challenges for pattern recognition and classification.
• Effective gene selection is crucial for identifying relevant biological markers and improving model performance.
• The curse of dimensionality hinders traditional classification methods in biological data analysis.

Purpose of the Study:

• To propose an effective gene selection method for high-dimensional microarray data.
• To enhance the performance of classification algorithms by reducing feature space.
• To address the challenges of dimensionality and small sample size in microarray data analysis.

Main Methods:

• Utilized information gain for initial feature relevance assessment.
• Employed a modified chaotic genetic algorithm (GA) with a chaotic mutation operator for enhanced population diversity and search capability.
• Implemented a K-nearest neighbor (KNN) classifier with leave-one-out cross-validation for performance evaluation.

Main Results:

• Successfully reduced the number of selected genes from high-dimensional microarray datasets.
• Achieved lower classification error rates compared to existing methods.
• Demonstrated the effectiveness of the enhanced chaotic GA in expanding search ability and improving classification accuracy.

Conclusions:

• The proposed information gain and chaotic genetic algorithm-based gene selection method is effective for microarray data.
• The enhanced chaotic GA improves population diversity, leading to better feature selection and classification performance.
• This approach offers a robust solution for the challenges in high-dimensional biological data classification.