SSE-BASED ABNORMAL POINT PROPORTION OPTIMIZATION METHOD AND DEVICE, AND COMPUTER DEVICE

摘要

Disclosed by the present application are an SSE-based abnormal point proportion optimization method and device, and a computer device. The method comprises: constructing a single classification support vector machine according to a current abnormal point proportion and a sample to be classified; classifying the sample to be classified to obtain a normal point center; obtaining the residual quadratic sum of each data point of an abnormal category and the normal point center so as to obtain a current residual quadratic sum; updating the current abnormal point proportion by subtracting step length from the current abnormal point proportion; classifying the sample to be classified according thereto so as to obtain a residual quadratic sum of each data point of the current abnormal category and the center of the normal point to serve as a next residual quadratic sum; dividing the difference between the next residual quadratic sum and the current residual error by the step length to obtain a residual error variation amplitude; and if the residual error variation amplitude exceeds a preset variation amplitude threshold value, adding the step length to the current abnormal point proportion to serve as the optimal abnormal point proportion.