Parkinson's syndrome aided prediction method based on multi-label model

Abstract

The invention relates to a Parkinson's syndrome aided prediction method based on a multi-label model. Prediction model building, balance evaluation and mining prediction stage processing modules are included. In the prediction model building module, a main syndrome is selected; K secondary syndromes are selected; the main syndrome and the K secondary syndromes form a new syndrome set; the balance of the syndrome set is evaluated; if the balance meets the conditions, the relevance between the main syndrome and the secondary syndromes is determined, and otherwise, the second step is carried out again; the syndrome set serves as training data to build the multi-label model. In the balance evaluation module, obtained syndrome sets are combined; the number and occurrence frequency of the syndrome combinations are counted; according to an information entropy formula, the entropy of the syndrome sets is calculated; the process is ended. In the prediction stage, the given to-be-predicted data are transmitted to the model to be classified; all the classification results of the model are voted, and a final prediction result is obtained. The method solves the problem that a multi-label classification algorithm is not accurate in prediction under sparse category and unbalanced conditions.

Description

technical field [0001] The invention relates to a modeling and prediction method for Parkinson's disease syndrome type, which solves the problem of inaccurate prediction of a multi-label classification algorithm under the condition of sparse and unbalanced categories. Background technique [0002] Data mining mainly refers to the non-trivial process of obtaining valid, novel, potentially useful, and ultimately understandable patterns from huge amounts of data. With the development of data mining, data mining technology has been widely used in various industries. [0003] The main objective of the present invention is to apply data mining technology to the diagnosis of Parkinson's disease in traditional Chinese medicine. In practice, it assists doctors in clinical diagnosis and treatment to better serve patients. [0004] Chinese medicine divides Parkinson's disease into five types of syndromes, each patient may be accompanied by one or two types of syndromes, and there are...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to solve the problem of unbalanced categories caused when the primary and secondary syndromes of Parkinson's disease are separated. sex

Images