A Hybrid Grouped-Artificial Bee Colony Optimization (G-ABC) Technique for Feature Selection and Mean-Variance Optimization for Rule Mining
Mrinalini Rana, Omdev Dahiya · International Journal of Engineering Trends and Technology · 2023
Data mining is a widely used method for analyzing and discovering knowledge in large data sets. It identifies the pattern hidden in the information by using mathematics and algorithms that can help discover hidden attributes. Data mining techniques are used to mine relevant patterns from large databases. It is broadly categorized into data preprocessing and mining results based on analysis outcomes. Soft computing procedures are used popularly these days for pattern predictions these days. This paper presents rule mining using Grouped - Artificial Bee Colony Optimization(G-ABC) technique for feature selection and mean-variance optimization for further rule mining. Classifiers are used to train and test the model for both feature selection and rule mining. For performing the experimental analysis of the work, Twitter and Baseball datasets were used. The proposed algorithm demonstrated the most optimized for the number of rules generated, the time required for calculation, and getting supplementary normalized information for rule mining. The best performer G-ABC with Neural Network (NN) classifier represents an average of 97.56% accuracy, a precision of 61.11, a recall of 96%, and an f-measure of 75% with G-ABC and mean-variance optimization technique with the Neural Network classifier.