Hi In page 21 of the GLM chapter (16), the notes say that for a simple linear regression the parameters can be estimated via matrix inversion using the following formula: \( {\beta} = (X^TX)^{-1}X^TY \) Is it correct to say that the \( (X^TX) \) is only being applied to make it a square matrix so that it will be invertible, and then the last \( X^T \) is then applied so that we're calculating \(X^{-1}\) and not "\(X^{-2}\)" ? thanks
The expression you've given is the MLE of the linear model so you can obtain it by maximising the log-likelihood.
