Active learning using rough fuzzy classifier for cancer prediction from microarray gene expression data.

Area of Science:

• Computational Biology
• Bioinformatics
• Machine Learning

Background:

• Cancer classification using microarray gene expression data is crucial but challenged by limited labeled samples.
• Traditional supervised methods struggle with accuracy due to insufficient labeled data.

Purpose of the Study:

• To propose a novel active learning method for enhanced cancer sample classification.
• To address uncertainty, overlappingness, and indiscernibility in gene expression data subtypes.

Main Methods:

• Developed an active learning method incorporating a rough-fuzzy classifier (ALRFC).
• ALRFC computationally selects informative samples to augment training datasets.
• Evaluated ALRFC on benchmark cancer gene expression datasets.

Main Results:

• The proposed ALRFC method demonstrated superior performance in cancer prediction.
• Achieved higher accuracy, precision, recall, F1-measures, and kappa compared to other techniques.
• Statistical significance confirmed the superiority of ALRFC via paired t-tests.

Conclusions:

• The novel ALRFC method offers an effective solution for cancer classification with gene expression data.
• ALRFC's ability to handle data complexities enhances predictive accuracy.
• This approach represents a significant advancement over existing active learning and supervised techniques.