Cancer prediction based on radical basis function neural network with particle swarm optimization.

Area of Science:

• Computational biology
• Artificial intelligence in medicine
• Biomedical data analysis

Background:

• Cancer poses a growing global health threat, often diagnosed at late, difficult-to-treat stages.
• Early cancer prediction is crucial for timely intervention and improved patient outcomes.
• Computational methods offer potential for proactive cancer risk assessment.

Purpose of the Study:

• To develop and evaluate a novel computational model for cancer prediction.
• To leverage machine learning for early detection of cancer.
• To enhance the accuracy and reliability of cancer prediction systems.

Main Methods:

• Utilizing a Radial Basis Function Neural Network (RBFNN) for cancer prediction.
• Optimizing the RBFNN using Particle Swarm Optimization (PSO).
• Training and testing the model on the Wisconsin breast cancer database.

Main Results:

• The optimized RBFNN demonstrated significant improvements in prediction accuracy.
• The method showed enhanced reliability and stability in cancer prediction outcomes.
• Experimental results confirmed the effectiveness of the PSO-optimized RBFNN.

Conclusions:

• The proposed PSO-optimized RBFNN is a highly effective tool for cancer prediction.
• This computational approach can aid in early cancer detection and prevention strategies.
• The method offers a promising advancement in applying AI to oncological diagnostics.