A clustering method of Chinese medicine prescriptions based on modified firefly algorithm.

Area of Science:

• Computational intelligence
• Bioinformatics
• Traditional Chinese Medicine

Background:

• Traditional K-means clustering algorithms exhibit limitations in analyzing complex datasets like Chinese Medicine (CM) prescriptions.
• These limitations include sensitivity to initial values and susceptibility to local optima, hindering accurate classification of CM medical cases.

Purpose of the Study:

• To develop and evaluate a novel, robust clustering method for Chinese Medicine (CM) medical cases.
• To address the shortcomings of traditional algorithms in handling the unique characteristics of CM prescription data.

Main Methods:

• A hybrid clustering approach integrating the Firefly Algorithm (FA) and Simulated Annealing (SA) was proposed.
• The method dynamically adjusts FA iterations and SA sampling based on fitness changes, enhancing swarm diversity and avoiding premature convergence.
• Expansion of the sudden jump scope in the swarm increases exploration capabilities.

Main Results:

• The proposed FA-SA hybrid algorithm demonstrated significant improvements over the traditional K-means algorithm in clustering CM medical cases.
• Experimental results showed enhanced individual diversity and superior clustering outcomes compared to K-means.
• The method provides valuable computational results for cluster analysis of CM prescriptions.

Conclusions:

• The collaborative Firefly Algorithm and Simulated Annealing method offers a more effective approach for clustering Chinese Medicine (CM) prescriptions.
• This advanced clustering technique overcomes the limitations of traditional methods, providing more reliable and diverse results.
• The findings suggest this method has significant reference value for the cluster analysis of CM prescriptions.