Moroccan industrial enterprises investing in the metallurgical, mechanical and electromechanical industries sector, play a significant role in Morocco’s economic development. The particularity of Moroccan companies investing in this sector is that the projects they are awarded generally require massive investment in the initial phase, and only generate sufficient cash flow in the long term to cover costs and debt servicing, especially when they operate for the benefit of public-sector customers, who are notorious for their late payments to suppliers. As a result, the loans granted have to be long-maturity, with longer grace periods than conventional loans. As a result, these projects do not have the same specific features as those of operating activities within other companies. In this study, some boosting machine learning models were used to predict the funding method of industrial projects. The data contains 5198 projects of some companies operating in the said sector, we found that K Neighbors Classifier, LGBM Classifier and Extra Trees Classifier can produce accurate and reasonably funding decision. Among the three machine learning methods, we find that ExtraTreesClassifier appears to be overall most effective.