Least squares loss L[ϕ]=Σt=1t(yi−f[xi,ϕ])2 Assumptions : independent, drawn from a Normal Distribution with mean μ=f[xi,ϕ]