Performance Evaluation of different Data Mining Techniques for different Medical Data Sets

摘要

Precision prediction has long been a complicated trouble with the existence of a big number of missing information in the dataset. Several designs offer with this specific challenge have both removed the missing information from the data collection as called as event deletion and applied some various ways to load that missing data. This report targets a story combining Prediction Product with missing information assertion to analyze several techniques and use Easy K-means clustering and apply the most effective one to an information set. The main goal of this report would be to propose a fresh cross multi-class SVM based on missing price imputation. The further advancement will be done by using information parallelism approach of similar computing.