Abstract
A very important problem in the area of neural networks and machine learning is the accurate evaluation of multidimensional integrals. An introduction to the theory of the stochastic approaches with a special choice of optimal generating vectors has been given. A new optimized lattice sequence with a special choice of the optimal generating vector have been applied to compute multidimensional integrals up to 100-dimensions. It is shown that the progress in the area of machine learning is strictly related to the progress in reliable algorithms for multidimensional integration.