Identification of Appropriate Features for Classification Using Clustering Algorithm

Now a day, developing the science and technology and technology tools, the ability of reviewing and saving the important data has been provided. It is needed to have knowledge for searching the data to reach the necessary useful results. Data mining is searching for big data sources automatically to find patterns and dependencies which are not done by simple statistical analysis. The scope is to study the predictive role and usage domain of data mining in medical science and suggesting a frame for creating, assessing and exploiting the data mining patterns in this field. As it has been found out from previous researches that assessing methods cannot be used to specify the data discrepancies, our suggestion is a new approach for assessing the data similarities to find out the relations between the variation in data and stability in selection. Therefore, Meta heuristic methods are used to choose the best and the stable algorithms among a set of algorithms.