question about BayesianRegression

[Keneyr]

why self.w_cov = np.linalg.inv(self.w_precision) in fit function? what's the meaning of w_cov? why y_var = 1 / self.beta + np.sum(x @ self.w_cov * x, axis=1) in predict function ?

Can anybody give me a clue or some knowledge about this? Thanks

[said-ml]

for the first question in statistic covariance matrix is inverse of precision matrix, as python code within in class (we access to argument through self) and using inv method from the package numpy.linalg we can write self.w_cov = np.linalg.inv(self.w_precision) in fit function because over all the because (at lest most of it the objective is to find out the weights )and the implementation direct of the result of the prdictive distribution page 156 equation (3-59) where here we take the basis function phi(X)=X