Аннотация: This scientific work focuses on the problem of storing and processing sensor data in modern microelectronics and their application in monitoring systems. An experiment was conducted using gas sensors as an illustration. The efficiency of calibration methods was examined using various machine learning algorithms to solve the problem of gas classification. In particular, the paper emphasizes the importance of model calibration in machine learning to enhance the accuracy of predictions and explain the linearity of the membership function.