Whale optimized mixed kernel function of support vector machine for colorectal cancer diagnosis

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Microarray analysis is crucial for colorectal cancer (CRC) classification.
• High dimensionality and imbalanced samples in microarray data reduce prediction accuracy.
• Accurate CRC prediction models are vital for clinical decision-making.

Purpose of the Study:

• To develop an ensemble model for accurate colorectal cancer (CRC) prediction using microarray data.
• To address challenges of high dimensionality and imbalanced samples in CRC datasets.
• To enhance the predictive performance of CRC classification models.

Main Methods:

• Feature selection using minimum redundancy maximum relevance (mRMR) to reduce dimensionality.
• Hybrid sampling algorithm RUSBoost to handle imbalanced data.
• Whale Optimization Algorithm (WOA) optimized mixed kernel function Support Vector Machine (MKF-SVM) for classification.

Main Results:

• The proposed ensemble model achieved superior G-means compared to existing models.
• Effective reduction in feature dimensionality and mitigation of imbalanced data issues.
• Demonstrated improved accuracy in classifying healthy and CRC patient samples.

Conclusions:

• The RUSBoost + WOA + MKF-SVM ensemble model significantly improves colorectal cancer (CRC) predictive performance.
• The model offers a robust solution for analyzing imbalanced microarray data in cancer research.
• This approach enhances the clinical utility of microarray data for CRC diagnosis and prognosis.