Hi guys,
I am trying to follow derivation 5.73 → 5.76 on Baysian linear regression.
Starting from the Book:
p(\omega|X, y) \propto exp\left(-\frac{1}{2}(y-Xw)^T(y-Xw)\right)\cdot exp\left(-\frac{1}{2}(\omega - \mu_0)^T\Lambda_0^{-1}(\omega-\mu_0)\right)
I expand to get this:
\propto exp\left(-\frac{1}{2}\left(\underline{-y^TX\omega-X^T\omega^Ty}+X^T\omega^TX\omega + \omega^T\Lambda_0^{-1}\omega - \underline{\mu_0^T\Lambda_0^{-1}\omega-\omega^T\Lambda_0^{-1}\mu_0}\right)\right)
But now the underlined terms are supposed to be
Read more… (47 words)