In the area of Machine Learning, classification is one of the important topics. Support Vector Machines (SVM) has thrived as a classification technique in that area in recent years and has been widely applied in many other areas. Here and now, we use it to classify new-thyroid in the area of Bio-medical. According to researches, SVM performs very well as a good classification tool when the effects of accuracy are compared among various classification algorithms. However, during the process of SVM operation, the accuracy of classification can be enhanced if training data can be processed beforehand in the phase of data pre-process. The main task of this research is to introduce Nonparametric Weighted Feature Extraction (NWFE) to pre-process the data and to use the data that have been processed to undergo SVM classification so that performing results are analytically compared between pre-process involved and that not involved. This research discovers that introducing NWFE is a way to be able to reduce the influence of noise so as to enhance the effectiveness of SVM classification. Hopefully, this research can evoke people’s attention to thyroid disease and then more precise classification techniques can be used in the area of Bio-medical in order to elevate people’s medical quality.
