Gene Optimized Association Rule Generation Based Integral Derivative Gradient Boost Classification for Disease Diagnosis

摘要

Associative classification is a significant technique used for disease diagnosis. Few research works has been developed for associative classification to predict the disease patients. However, the performance of conventional associative classification technique was not efficient. In order solve this limitation, A Gene Optimized Association Rule Generation based Integral Derivative Gradient Boost Classification (GOARG-IDGBC) technique is proposed. The GOARG-IDGBC technique is designed for diagnosing the disease with minimal time consumption and higher classification accuracy. Initially GOARG-IDGBC technique used Optimized Genetic Algorithm (OGA) to generate the association rules from attributes in a medical dataset using support and confidence value. By using generated association rules, classification is then carried out using Integral Derivative Gradient Boost Classifier (IDGBC) in GOARG-IDGBC technique. Integral Derivative Gradient Boost Classifier classifies the patients in a medical dataset as normal or abnormal with higher classification accuracy through constructing strong classifier. Experimental evaluation of GOARG-IDGBC technique is carried out on factors such as classification accuracy, disease diagnosing time and false positive rate with respect to different number of patients. The experimental results show that the GOARG-IDGBC technique is able to improve the classification accuracy and also minimizes the time of disease diagnosing when compared to state-of-the-art works.