Logic mining classification for Phone Prices dataset using Discrete Hopfield Neural Network and Weighted Random 2 Satisfiability

Smartphones have become essential in today’s technology-driven world, with various models offering unique features like camera quality, screen resolution, and storage. Understanding how these features influence smartphone prices can help consumers make informed purchasing decisions. This study introduces a logic mining technique to classify smartphone features that contribute to pricing using Weighted Random k Satisfiability with Modified Reverse Analysis. The model implements a Discrete Hopfield Neural Network, a Modified Niched Genetic Algorithm for training, and the Jaccard Feature Selection Method. The Phone Prices Dataset from Kaggle was used for experimentation, revealing the model’s ability to extract optimal patterns in the form of induced logic. The results show that the proposed model outperforms existing methods, achieving an accuracy of 0.8083, precision of 0.8925, specificity of 0.9760, Matthew’s correlation coefficient of 0.5334, and an F1-score of 0.5887, demonstrating its effectiveness in analyzing and classifying smartphone pricing factors.

Text 125485.pdf - Published Version Available under License Creative Commons Attribution Share Alike. Download (561kB) Official URL or Download Paper: https://ojs3.unpatti.ac.id/index.php/barekeng/arti...

Actions (login required)

View Item