PROBABILISTIC TRAINING FOR BINARY NEURAL NETWORKS

摘要

A method (100) for training an artificial neural network (1) that is at least partially implemented as a binary neural network (2), comprising• receiving (110) training input values (11) for the inputs x_i of the artificial neural network (1),• processing (120) training input values (11) to obtain output values (12),• applying (130), to obtained output values (12), a loss function (13);• updating (140) training weights w_lj and branching back (150) to processing (120) until rating (13a) by the loss function (13) meets a predetermined termination criterion (160); and• configuring (170) actual artificial neural network (1, 2) according to finally obtained training weights w_lj,wherein processing (120) comprises:• determining (121), for each neuron I in the layer (21-23), a distribution A_l of pre-activations a_l using a distribution W_lj of weights w_lj; and• determining (122) a distribution H_l of output values h_l by applying, to distribution A_l, predetermined thresholding function g.