Applicability of convoluted neural networks to the dataset fitting problem

We analyzed the applicability of convolutional neural networks to solving the dataset fitting problem. The convolutional neural network was trained with training datasets containing function curves. We selected 6 linearizable functions. The convolutional neural network detected the functional relations in the datasets taken from the MNIST database intended for statistical software testing. The results show that it is possible to use the proposed approach for visual correlation analysis and curve-based data fitting.