I am having this simple confusion.Lets consider a multivariate gaussian distribution with mean $\mu$ and precision matrix $K$. Then the exponential term is
$$(x-\mu)' K (x-\mu)$$
If I open the above expression, I will get $x'Kx- x'K\mu-\mu'Kx + \mu'K\mu$. The middle two terms are not equal, they were supposed to be same.
So I am wondering if I am doing something wrong. Any suggestions?
The two "middle terms" are each other's transposes. Generally a matrix is not equal to its transpose. But notice that these are $1\times 1$ matrices, and every one of those is equal to its transpose.