Neoadjuvant Chemotherapy In Breast Cancer Prognosis State Prediction System Developmentbased On Data Mining

Breast cancer is a malignant cancer that leads to a high death rates.In recent years,neoadjuvant chemotherapy has been taken to the cure of breast cancer and achieved pretty good performance.The emergence of neoadjuvant chemotherapy is a good news to females as it decreases the chances of breast excision as well as the death rates.Neoadjuvant chemotherapy brings effective treatment for most patients,however,for minor percent of patients the treatment is not satisfying.Thus the prediction of treatment becomes a significant matter.In the past this process is taken by the experienced doctors,yet the decision is subjective and the qualification for doctors is demanding.Data mining is quiet suitable tool to discover valuable information from enormous data and help making decisions.This paper makes use of data mining technique to develop a prediction system for neoadjuvant chemotherapy.This paper first take a preprocessing for data and the adopt decision tree and Support Vector Machine(SVM)to perform a prediction.This system can help the doctors to make decision and choose the proper treatment scheme when take neoadjuvant chemotherapy.The experiments of this paper indicate that SVM technique obtains a better performance,and the adoption of different kernel functions accounts for the final results.In this paper,we compare the polynomial kernel function and radial basis function and find that radial basis function achieves better results.Therefore,this paper chooses SVM method with radial basis function as the kernel function to develop the system.The system can be an aided prediction software system for doctors to perform neoadjuvant chemotherapy.