Pre-processing in data mining played essential role for enhancing data quality. The basic concept behind is that, learning with accurate and high quality data may provide more efficient classification results as compared to learning with poor quality of data. In this presented paper a pre-processing technique is implemented with slight modification which is based on the technique given in [1]. In this paper a promising approach of data pre-processing is provided which utilizes a fuzzy technique in order to improve the data quality. The implementation of available technique is performed using MATLAB. Additionally, the improved fuzzy technique is also implemented with it. The results demonstrate the effectiveness in classification accuracy after implementation of both techniques. Finally, the obtained results favour the proposed model for enhancing the performance of classifiers in both manners supervised and unsupervised manner.

Data mining pre-processing data quality enhancement classification performance improvement