Application of a Stacking Algorithm to Identify Deviations in Gas Transmission Network Data

Machine-learning algorithms are increasingly being used to solve problems of identifying deviations in data. However, the accuracy of the algorithms varies, and not all algorithms are equally useful. In a number of problems, it is advisable to use ensembles of algorithms, among which a special place is occupied by the stacking algorithm, the peculiarity of which is that it can use algorithms of different types. The article discusses the application of the stacking algorithm in comparison with the LSTM neural network.