Optimal feature combination-based re-hospitalization behavior prediction system and method

Abstract

The invention discloses a rehospitalization behavior prediction system and method based on an optimal feature combination, which processes and analyzes long-term hospitalization records, general outpatient records, chronic disease records and physical examination records, and obtains multiple data that can represent the patient's rehospitalization environment. Rehospitalization characteristics; design a reasonable fitness function, use genetic algorithm to obtain the optimal feature combination that affects rehospitalization; use the optimal feature combination and rehospitalization label as the input of the prediction method, and use the integrated learning algorithm to judge and predict whether the patient is Performing rehospitalization within the observation window can predict whether the patient will be rehospitalized within a certain time interval, accurately express the patient's rehospitalization environment, and help patients themselves and medical institutions to make reasonable arrangements for time, medical care, and financial resources.

Description

technical field [0001] The invention relates to a rehospitalization behavior prediction system and method based on an optimal feature combination. Background technique [0002] Health is an inevitable requirement for promoting the all-round development of human beings and a basic condition for economic and social development. Disease prevention and early warning work can effectively improve people's health and reduce medical expenses. Under the environment where the government and the whole people pay attention to health, the accurate prediction of rehospitalization has gradually attracted the attention of scholars and the industry. With the continuous development of information technology, the field of health care has gradually accumulated a large amount of data, forming big data of health care. Big data technology can help researchers extract data related to disease prevention and disease early warning from complex and heterogeneous medical data, including long-term medi...

AI Technical Summary

Problems solved by technology

[0003] The current research direction is mostly aimed at the pathology, disease prevention and disease early warning of patients, but there are few evaluations and predictions on whether patients with existing diseases are hospitalized, which will lead to possible funding or Improper time arrangement will cause conflicts and bring health or economic losses; at

Images