Machine Learning with Swift

Normal equation for multiple linear regression

If you want to implement regression in a production code, don't use matrix inversion operation explicitly. The problem with it is that it is very numerically inefficient. Instead, you can use one of the functions for solving a system of linear equations, which is essentially the same as finding regression coefficients. The good fit from the LAPACK package (part of the Accelerate framework) is the QR-decomposition function.