Breast cancer diagnosis through Data Mining techniques

Abstract

The cancer is a very critical disease is affected from both men and women. It s newlinedifferent types and affected different place in human body. The breast cancer mostly newlineaffects in women. It has so many diagnosis systems available in our country. But it has newlinesome problems to solve the breast cancer diagnosis methods. It has so many factors, the newlinebreast cancer data is not fulfilled, inconsistent not formatted. newlineIt should be processed is not getting accuracy for diagnosing problems. I am going to newlineapply the different data mining techniques and tools to solve the breast cancer diagnosis newlineproblem. The breast cancer has different data sets. The first one is breast cancer tissues data newlineand another one is prognosis data. But the last one is diagnosis of breast cancer data. newlineIt has some problems should be occur from the breast cancer diagnosis dataset. The newlinefirst one is very large size of the dataset and the second one is different range of data. In our newlineresearch to solve the above problems using a breast cancer diagnosis dataset. It has to be newlineimproved the classification accuracy and reduces the processing time. A number of newlineresearchers have used to various data mining techniques and tools to address the breast newlinecancer diagnosis problems. But it s not improves the classification accuracy and reduce the newlineprocessing time. It has some defect should be occur from the diagnosis of breast cancer newlineproblem. newlineIn this problem to solve a, very common data mining algorithms deployed include newlineDecision Trees (DT), Neural Networks (NN), Naive Bayes (NB), Support Vector Machine newline(SVM) and K-Nearest Neighbor (K-NN) algorithms and among others for feature selection newlinemethods such as Info Gain, Fisher Score, Relief and feature extraction method for PCA. newlineThese techniques are applied in classification for both qualitative and quantitative breast newlinecancer diagnosis data and the interpretation of the classification models was created. Some newlineof these techniques do not give any assurance on the precision of the constructed models in newlinelarge data and accuracy. But my propos