Predicting Protein in Cancer Diagnosis using Effective Classification and Feature Selection Technique

Cancer is the ailment where some cells in the body develop abnormally and wreck the opposite surrounding cells and their normal functions. Most cancers spread throughout the human body. Prediction of the protein using genes dataset causing cancer is of terrific significance in cancer prognosis and discovery of the drug. The DNA microarray approach concurrently monitor hundreds of gene expression viable. Using this massive quantity of Genes information, the prediction of protein causing cancer is explored. In this paper diverse information mining strategies are used that are expected to predict the protein causing cancer. The first stage of retrieving the records is characteristic extraction where Brute Force Algorithm is used in which the crucial features are extracted from the genes dataset which improvises the device performance, then the Fuzzy C Means the clustering approach is used to divide the gene dataset into numerous clusters while predicting the new patients the machine verify to which cluster it belong and examine the facts most effective with records of that precise cluster therefore the performance of the system is improvised and KNN algorithm is used for predicting protein. The protein prediction might help the clinical sector in the diagnosis of most cancer and additionally help to discover the drug for the protein causing cancer.