Program for computer simulation of X-ray absorption spectra of materials based on time series prediction and machine learning algorithms
The program is designed to create datasets simulating X-ray absorption spectra using time series prediction models and machine learning algorithms. The initial data are: X-ray absorption spectra obtained during the experiments and presented in the form of two-dimensional flat tables. The program implements machine learning algorithms and artificial neural networks: RNN2Dense (recurrent fully connected artificial neural network); Seq2Seq with LSTM cells; Seq2Seq with GRU cells. The program generates time series based on machine learning algorithms, saves and visualizes data. The properties of the predicted time series are analogues of X-ray absorption spectra. The program can be used to create data samples when solving problems of classification of materials, analysis of experimental data of X-ray absorption spectra.