Stochastic Approach for System Identification using Machine Learning
Shokhjakhon Abdufattokhov, Behzod Muhiddinov · 2019
This paper gives a clear understanding of how a system can be modeled in a non-parametric and probabilistic way using machine learning based on historical data with taking into account uncertain noise added to the output of the system. We explained key concepts step by step and end up with a unique algorithm. Moreover, we demonstrated an illustrative example of the proposed algorithm in case of nonlinear dynamic system.