MACHINE LEARNING PROBLEM

I have a data frame called Project for quantities of machine parts in stock . It has 10 variables(columns), unit price, criticity( importancy of the part, its 0/1 for yes/no), the type of the part (mechanical, electrical,...its onehot encoded), quantity in stock, installed quantity, etc...

we wanna use these variables to predict the best quantity the company should keep in stock because they usually have an excess.

I am doing this in Rstudio, we should use a machine learning algorithm to predict these quantities. how can i use random forests to predict in the best way possible? also what if there is a way other than random forests, that is better to predict?