Research On On-Line Medical Pre-Diagnosis Classification Model And Algorithm With Privacy Protection

The on-line medical pre-diagnosis platform is an intelligent system that automatically gives a preliminary diagnosis conclusion based on the symptoms described by the user.It can help medical users to carry out self-diagnosis,especially after the outbreak of the COVID-19,an on-line platform that can provide medical prediagnosis services anytime,anywhere is particularly important for people who are unable to go out for medical treatment in time.A satisfactory online medical prediagnosis platform needs to meet the practical requirements of privacy-preserving,high precision and high efficiency at the same time.The third is the selection of the sample generation method for the characteristics of large volume of medical data.And it is necessary to overcome the impact of the explosive growth of medical data and imbalance on the efficiency and accuracy of the pre-diagnosis.In view of the above problems,this paper studies the classification model and algorithm of on-line medical pre-diagnosis with privacy-preserving.We build an online medical pre-diagnosis service framework with privacy protection.The main research work is as follows:Firstly,aiming at the privacy leakage problem in the on-line medical pre-diagnosis platform,this paper proposes an on-line medical pre-diagnosis classification model(OMPD)with comprehensive privacy-preserving.We preprocess medical data such as feature weight assignment,iterative transformation,forward expansion and lightweight encryption,and use signature technology for interactive information.The cloud uses the Relief-k MW classifier for pre-diagnostic classification.At the same time,the cloud can provide relatively accurate pre-diagnosis results without obtaining the original medical data.Both theoretical proof analysis and experimental results demonstrate the effectiveness of the pre-diagnostic classification model in this paper.Secondly,aiming at the problem that the performance of the on-line medical prediagnosis classification model is easily affected by the large and imbalanced medical data,this paper proposes a learning sample generation algorithm(W-LVQ)based on incremental learning vector quantization.The algorithm generates medical samples by learning and compressing real medical data.These medical samples are not only small in size and well-balanced in distribution,but also protect sensitive information in medical data in a way that does not expose the original data.The experimental results verify the feasibility and effectiveness of this method.