Application to Finance

We have the code which was uploaded to GitHub. And so it's just sufficient to take it and to run it. So it's very fast. It literally replaced calls to neural networks without any adaptations. But at the same time, I would like to mention two errors of the application of our methods. The first one is a tabulation or approximation of a slow function. When we can spend some time on the calculation of the slow function, then we fit it using our methods. And then we just price it, calculate it instantaneously.