Group Method of Data Handling(GMDH) is an algorithm that aids in the development and training of polynomial neural network (PNN)-based predictive models. PNN models have the advantage of being able to self-select the number of inputs and topology required to optimally and reliably learn the presented problem.
In optimization problems, the Group Method of Data Handling, or GMDH, can be used as an optimization technique. Because the method is fast, non-parametric, and cognitive, the identification of the optimal point will be faster, and the result will be more reliable than that of non-linear or simple neural network models.
A tutorial is created to describe how to use GMDH in optimization problems with the help of the GMDH Shell software.
Happy optimizing.
Comments