The words contained in this file might help you see if this file matches what you are looking for:
...The matrix cookbook a notation vector always denotes row with at we will d denote column transpose t b ab product xa ij ik ki k c bc ba inner of vectors x derivatives gradient function f r is dened as chain rule z y yz linear ax quadratic figure regression training data points are given by blue circles original plotted green line based on knowledge from want to nd wearegiven dataset hx i for simplicity assume that scalar n p dimensionality w which minimizes error e yi setting illustrated in can be easily written form noting sum over squared terms represented xw yn and nz least squares solution wenowderivate set apply eq polynomial basis functions wtxtx ytx rightmultiply isalso known pseudo inverse instead directly using input our use arbitrary non tw l degree have higher dimensional take all products elements up order thus number quickly increases radial here uniformly distribute also called receptive elds space exp...