Editorial: Special Issue on Machine Learning Methods for Inverse Problems

Abstract

The use of Machine Learning methods has drastically increased in recent years. This mainly obeys to the fact that we now have enough data to extract insights from it, including but not limited to, parameter estimation/calibration, forecasting, and prediction. Besides, computers have evolved in such a manner that computational-intensive statistical methods can be applied in practice. This special issue presents some recent advances in the use of Machine Learning methods in Inverse Problems. These problems can be found in many branches of science. Typically, one wants to tune or estimate model parameters to mimic (or to explain) real-data behavior. The problem reduces to, given some datasets (and depending on the chosen model), what parameters (states) better describe their behavior.