Logic mining method via hybrid discrete hopfield neural network

The growing interest in logic mining as a knowledge extraction tool in data mining has attracted considerable attention from researchers. Despite the success, the limitations of existing logic mining methods are often overlooked, hindering the search for optimal solutions in binary classification tasks. To address these challenges, this paper introduces a novel logic mining approach using the Y-type Random 2 Satisfiability logical rule, combined with hybrid mechanisms within the Discrete Hopfield Neural Network. The first contribution involves the incorporation of a Hybrid Differential Evolution Algorithm to accelerate the optimization of synaptic weights during the training phase. Additionally, the retrieval phase is enhanced by proposing a swarm mutation operator, which diversifies the final neuron states, thereby broadening the solution space. Furthermore, an improved reverse analysis method is applied to optimize attribute selection and generate the most effective training logic. To demonstrate the efficacy of the proposed logic mining approach, experiments were conducted using both simulated and real-world datasets. The results indicate that the proposed model significantly outperforms baseline models across all performance metrics. The study concludes that the enhanced logic mining technique effectively captures the knowledge of datasets and facilitates transparent decision-making, making it a valuable tool for both researchers and practitioners.