Hepatic encephalopathy prediction model and method based on diffusion enhanced feature selection neural network

The invention provides a hepatic encephalopathy prediction model and method based on a diffusion enhanced feature selection neural network, and the method comprises the following steps: obtaining sample data, and dividing the sample data into a training set and a test set; constructing a data enhancement module by using a diffusion model, and inputting the training set into the data enhancement module to obtain a training set after sample expansion; constructing a judicial feature interaction module, and inputting the expanded training set into the judicial feature interaction module to obtain data after feature interaction; constructing a judicial feature selection module, taking the data after feature interaction as the input of the judicial feature selection module, and obtaining the final output after feature selection; and constructing a hepatic encephalopathy prediction model of the unique feature processing neural network, and training the model. Through the diffusion model data enhancement technology,