Bayesian network inference method based on random computing theory

Abstract

The invention relates to an inference method, in particular to a Bayesian network inference method based on a random computing theory. According to the method, the calculation process of the Gaussianrandom generator based on the central limit theorem is simplified; the multiplication operation in the Bayesian neural network reasoning process is simplified; a 0-1 sequence obtained by using a binomial distribution generator is used as a data unit, multiplication operation is realized by using an AND gate, and a 0-1 random number sequence directly participates in calculation to complete Bayesianneural network inference, so that the purposes of reducing hardware resources, improving the calculation speed and reducing the system power consumption are achieved.

Description

technical field [0001] The invention relates to a reasoning method, in particular to a Bayesian network reasoning method based on stochastic calculation theory. Background technique [0002] In recent years, the growth of data scale and computing power has greatly promoted the development of artificial intelligence. As an active branch of artificial intelligence, deep learning has made continuous progress. At present, many efficient neural network models have been proposed by industry and academia, which are different from deep neural network models. Determining the weight value of the model, the Bayesian neural network parameters usually conform to a fixed probability distribution, assuming that the weight ω conforms to a Gaussian distribution with a mean of μ and a variance of σ, in the reasoning process, first obtain a random number from the standard normal distribution h, and then use the linearity of the normal distribution, ω=σh+μ to obtain the final weight value, when...

AI Technical Summary

Problems solved by technology

In this process, all the data are numerical data, and the operation adopts the traditional multiplication and addition calculation, which will consume a lot of hardware resources, and it takes multiple cycles to generate a random number, the calculation speed is slow, and the system power consumption is high.

Images