High Utility Item Set Mining in Transactional Data using a Deep Learning Model with Particle Swarm Optimization

Nowadays, when it comes to big data, finding highutility itemsets from transactional data aids in decision-making and insight development. The conventional methods used in highutility itemset mining often face issues of scalability and efficiency due to the large volume of data and complex computations. This article presents a new approach to enhance the accuracy and speed of high-utility itemset mining using an optimization algorithm and a deep learning model. Specifically, a hybrid model that integrates a deep neural network (DNN) with an optimization algorithm, namely Particle Swarm Optimization (PSO), is proposed to enhance utility measures and improve the mining process. The DNN learns complex relationships between items in transactions, while PSO optimizes the itemset selection process, balancing computational time with utility improvement. Simulations of the model on different datasets using MATLAB demonstrated high accuracy, precision, and mean absolute error (MAE) of 97.94%, 97.04%, and 43.72%, respectively. This further confirms the effectiveness and real-time applicability of the proposed method in practical scenarios.