Automatical Optimal Threshold Searching Algorithm Based on Bhattacharyya Distance and Support Vector Machine

Medical activities produce a huge amounts of medical data which include a large number of redundant data. Feature selection is an effective method to remove the redundant data. Selecting the feature attributes needs to determine a threshold. Currently, the thresholding value is mainly set by one´s determination manually, which may lead to inaccurate threshold. In this paper proposes forward an adaptive threshold method combining Bhattacharyya distance and support vector machines (SVM). The experimental results show that the proposed method not only makes a high credibility threshold, but also improves classification work efficiency.