文献类型：专利

英文题名：Unbalanced data classification system boundary-based digestion pseudo inverse algorithm, has calculating testing sample point based on probability of candidate sample, and determining final decision to obtain current testing sample point

作者：WANG Z;LI D;ZHU Y;GAO D

机构：[1]UNIV EAST CHINA SCI & TECHNOLOGY

申请号：CN107092927-A

申请日：2017-04-01

公开日：2017-08-25

语种：英文

收录：DERWENT

摘要：NOVELTY - The algorithm has a set of instructions for transmitting collected sample into a vector model. Training process is performed to obtain three classification decision surfaces. A classification strategy is obtained based on pseudo inverse algorithm to train a sample point and classify a decision surface. Two distance vectors are obtained to obtain sample point distance of a candidate and decision distance. A testing sample point is calculated in a testing stage based on probability of a candidate sample. A final decision is determined to obtain a current testing sample point. USE - Unbalanced data classification system boundary-based digestion pseudo inverse algorithm. ADVANTAGE - The method enables improving classification accuracy and greatly shortening debugging time. DESCRIPTION OF DRAWING(S) - The drawing shows a flowchart illustrating an unbalanced data classification system boundary-based digestion pseudo inverse algorithm. '(Drawing includes non-English language text)'