A miscellaneous conclusion & some limitations for the hklam statistical model theory

There may be numerous discussions about my HK-Lam Statistical Model theory. In practice, the theory is about turning any giving random matrix into a linear combination through some suitable linear transformation. Through the linear combination, we may then convert it into the linear regression (without reality engineering error term). In order to fill the gap between pure theory and the apply engineering, one may then use the machine learning (e.g. gradient descent) method to find out the expected error term. At the same time, for the inverse of the HKLam Statistical Model Theory, there may be a 1/x → infinity situation, this author thus proposes a primal duality gap (delta > 0, = 0, < 0) cases to solve such inverse problem (or alternatively, in the mathematical geometric sense, by definition, the inverse of the linear transformation is just a constant k multiplied with a normal unit vector i.e. x-1 = (x / ||x||2) = (1/||x||) * (x/||x||)=k*n = and k = (1/||x||) and ||x|| NOT equal to zero.Otherwise, we have the fact that A0 = 0 is the only trivial solution and thus the unit normal unit vector will definitively NOT equal or even exclude the case of a zero vector.  In fact, k = ||A (x1, x2, …, xn)-1(b1, b2, …., bn)-1||. Hence, from the inverse HK-Lam theory, we may find some more hidden social discussing issues or implications from the final Nash Equilibrium matrix and thus we may further make those essential and necessary improvements, policy and contributions to our society.