Missing medical diagnosis data imputation method and apparatus, electronic device and medium

An electronic device, comprising:

• one or more processors;
• a memory, configured to store one or more programs;
• the one or more programs, when executed by the one or more processors, cause the one or more processors to implement a method comprising:
• acquiring raw data with data missing, wherein the raw data is a medical diagnosis dataset with the data missing;
• randomly dividing the original data into initial sample point data and candidate sample point data, and constructing and training a generative adversarial network initial imputation model by utilizing the initial sample point data;
• estimating a change in a parameter of the generative adversarial network initial imputation model by sample points in the candidate sample point data;
• on the basis of the change in the parameter of the model, calculating an influence of the sample points in the candidate sample point data on a prediction result of the generative adversarial network initial imputation model by utilizing a chain rule;
• estimating the prediction result of the adversarial network initial imputation model by utilizing the influence;
• sampling a sample point with highest influence in the candidate sample point data by utilizing a binary search algorithm, and further iteratively optimizing the generative adversarial network initial imputation model to obtain a generative adversarial network imputation model;
• and
• imputing missing data for the medical diagnosis data by utilizing the generative adversarial network imputation model.

A non-transitory computer-readable storage medium storing a computer program, wherein the program, when executed by a processor, implements a method comprising:

• acquiring raw data with data missing, wherein the raw data is a medical diagnosis dataset with the data missing;
• randomly dividing the original data into initial sample point data and candidate sample point data, and constructing and training a generative adversarial network initial imputation model by utilizing the initial sample point data;
• estimating a change in a parameter of the generative adversarial network initial imputation model by sample points in the candidate sample point data;
• on the basis of the change in the parameter of the model, calculating an influence of the sample points in the candidate sample point data on a prediction result of the generative adversarial network initial imputation model by utilizing a chain rule;
• estimating the prediction result of the adversarial network initial imputation model by utilizing the influence;
• sampling a sample point with highest influence in the candidate sample point data by utilizing a binary search algorithm, and further iteratively optimizing the generative adversarial network initial imputation model to obtain a generative adversarial network imputation model;
• and
• imputing missing data for the medical diagnosis data to be imputed by utilizing the generative adversarial network imputation model.