Our study focuses on the development of new Estimation of Distribution Algorithms (EDAs) with neuro-evolution for pseudo-Boolean optimization problems. We define a strategy for updating the frequency vector at each generation using a neural network, trained by an evolutionary algorithm. To evaluate the effectiveness of our approach, we carried out experiments on instances of the Quadratic Binary Unconstrained Optimization (QUBO) problem of different types. The algorithm automatically discovered demonstrates its competitiveness with existing EDAs in the literature.