This document presents an intelligent visualization framework for multi-dimensional data sets. The framework includes pre-processing, feature selection, classification, rule refinement, and visualization phases. In the feature selection phase, principal component analysis and rough sets are used to select important features. Classification is done using rough set rules generation. The rules are then refined using entropy and genetic algorithms. Finally, the refined rules and reducts are visualized using nodes, edges, charts and grids to help experts understand the data. Experimental results on breast cancer and prostate cancer data sets demonstrate the performance of the approach.