Band gap is an important and key parameter in semiconductors for electronic and optical applications. Despite the many advances in computing of first to the resident, accurate calculation of the band gap is still a challenging issue and often requires heavy calculations.On the other hand, machine learning methods along with existing large databases have led to the production of simpler solutions to many complex problems.Therefore, in this project, we will use machine learning methods to produce a suitable model for predicting the band gap of semiconductor crystals.For this purpose, we first select a set consisting of double semiconductor crystals and then, by computing of first to the resident, we prepare the required data.Then, using MATLAB software and LASSO, OLSR, Ridge, SVR methods, we will pay to Supervised Learning of machine.