The case deals with the ECG signal analysis using logistic regression. One of the most pressing cardiology challenges today is to improve the quality of automated electrocardiogram (ECG) analysis. Efficient algorithms allow physicians to use long-duration recordings in order to obtain important diagnostic information about a patient's cardiac activity. Contestants solve the task of developing an algorithm, which classifies the QRS complexes into "pathological" and "healthy" categories using the method of logistic regression. The input data include ECG recordings, as well as QRS complexes positions already marked. The task is to identify features for classification, construct and optimize the model, and evaluate its sensitivity and predictive power. Register and get personal access to cases on the Moodle platform.