Accepted_test

The work is devoted to building an explainable machine learning model to predict the risk of all-cause mortality in patients with type 2 diabetes mellitus (T2DM). To build the model, a dataset containing 568 patients with type 2 diabetes and more than 100 general, clinical and laboratory characteristics (features) was used. Various survival models were considered in conjunction with various strategies for selecting an optimal subset of features. The resulting model, which is based on a random survival forest algorithm, demonstrates high predictive accuracy over various prediction time horizons.