Medical project identification method and device, equipment and computer readable storage medium

Abstract

The invention relates to machine learning, and provides a medical project identification method and device, equipment and a medium. According to the method, the average value and the standard deviation of the medical project cost are extracted as the characteristic data, the characteristic data are divided into the measurement-free characteristic data and the target characteristic data, and finally the target characteristic data are subjected to anomaly identification based on the overall characteristic data through the isolated forest model, so that the model does not need to evaluate the careless degree of all the characteristic data and calculated quantity of the model is reduced; compared with a traditional mode of completely screening key medical projects according to experience, themethod has the advantages that human resources are saved to a great extent; compared with a medical project identification mode based on other dimension reduction algorithms or feature selection algorithms, the isolated forest algorithm is used as a full-automatic algorithm, parameter adjustment is not needed, a large amount of computing power is saved, and therefore the identification effect of existing medical project key identification is improved in a multidirectional mode. In addition, the invention also relates to a blockchain technology, and the feature data can be stored in the blockchain.

Description

technical field [0001] The present invention relates to the technical field of machine learning, in particular to a medical item identification method, device, equipment and computer-readable storage medium. Background technique [0002] In the real-world data set, the number of medical items is very large, even in diseases with relatively clear treatment plans, it can reach as many as thousands. In various applications such as medical insurance cost control and medical pattern recognition, managers usually It is hoped that some key medical projects can be screened out for focused analysis and research. However, this step currently faces at least the following difficulties: most decision makers cannot grasp the key medical items of all diseases; pure manual identification takes a lot of time; the existing machine learning dimensionality reduction method takes a long time to calculate; the existing dimensionality reduction method Poor interpretability. Therefore, the above-...

AI Technical Summary

Problems solved by technology

[0003] The main purpose of the present invention is to provide a medical item identification method, equipment and computer-readable storage medium, aiming to solve the technical problem of poor identification effect of existing medical item key identification methods

Images