Abstract
We proposed that polaritons are approximately governed by Schrodinger equation, and one of the powerful method are perturbation series expansion when our quantum neural nets. And an explicit expression of solution for quantum neural nets is given by applying the convergence type of both initial and final state vectors. Then I attempted to rewrite the classical neural network into quantum neural net, and we noticed the similarities with both classical network and quantum one as for network construction and the method of weight calculation. However, quantum neural net belongs to the world of complex numbers, and its results of calculations do not mean probability but probability amplitude. So, the most different between quantum net and classical one exists in whether those neural nets have quantum interferences or not. We clearly obtained the concrete method of calculations and quantum Bayes' form by using state vectors.
