A method of tumor classification based on wavelet packet transforms and neighborhood rough set

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Tumor classification from gene expression data presents challenges due to high dimensionality and small sample size (SSS).
• Identifying relevant genes is crucial for accurate tumor classification, necessitating effective feature extraction and selection methods.

Purpose of the Study:

• To propose a novel tumor classification approach integrating Wavelet Packet Transforms (WPT) and Neighborhood Rough Sets (NRS).
• To enhance the efficiency and accuracy of gene expression data analysis for tumor classification.

Main Methods:

• Feature extraction was performed using Wavelet Packet Transforms (WPT).
• Decision tables were constructed, and attribute reduction was achieved using Neighborhood Rough Sets (NRS).
• A refined feature subset with high classification ability was obtained.

Main Results:

• The proposed WPT and NRS method effectively extracts relevant features from gene expression data.
• Attribute reduction by NRS significantly improved the classification ability of the selected feature subset.
• Experimental validation on three gene expression datasets confirmed the method's effectiveness and feasibility.

Conclusions:

• The combined WPT and NRS approach offers a powerful tool for tumor classification using gene expression data.
• This method addresses the challenges of high dimensionality and SSS in genomic datasets.
• The approach demonstrates significant potential for improving diagnostic accuracy in cancer research.