In this work, logistic regression was applied within the Python programming language to various data sets. Binary logistic regression and multicategory logistic regression are also described. The learning process of the model is explained with reference to the implementation in the Python programming language, and the basic meaning of the parameters. One of the key aspects of the work is the analysis of model accuracy based on input data. For this purpose, the basic characteristics used to describe the accuracy of the model are explained, such as the matrix of confusion, precision, sensitivity and ROC-curve. These metrics allow evaluating the performance of logistic regression in problem detection and classification. Furthermore, logistic regression was compared withotheralgorithmsusedinmachinelearning. Thiscomparisonaimstoshowtheadvantagesand disadvantages of logistic regression compared to other algorithms, and to point out the situations in which logistic regression is most suitable for application in engineering problems. In the conclusion of the paper, it is emphasized that logistic regression is a very powerful tool for engineers. Its application can help solve various engineering problems, such as predicting failures in a water supply system.