<![CDATA[This research focuses on optimizing dataset classification by combining hybrid grid partitioning and rough set-based fuzzy rule generation. Traditional classification algorithms often face challenges in handling high-dimensional data, attribute redundancy, and uncertainty, leading to reduced accuracy and increased computational complexity. To address these issues, we propose an integrated approach that leverages hybrid grid partitioning for adaptive representation of the dataset, rough set-based attribute reduction for identifying relevant attributes, and fuzzy rule generation for handling uncertainty and capturing complex relationships. The hybrid grid partitioning creates multiple levels of granularity, capturing both local and global patterns. Rough set-based attribute reduction reduces dimensionality and eliminates redundant information. Fuzzy rule generation enables the handling of uncertainty and the mapping of input attributes to output classes. We present a case example of customer churn prediction in a telecommunications company to illustrate the practical relevance of the proposed approach. The optimized classification model provides insights into churn factors and enables proactive measures for customer retention. The research contributes to the field by offering an integrated framework to enhance classification accuracy, interpretability, and efficiency. The findings have the potential to benefit various industries and applications that rely on accurate classification models, improving decision-making processes and performance of intelligent systems.]]>
Copyrights © 2020
