Neural networks or deep learning have become synonymous with artificial intelligence (AI), and the idea that human intellectual work will soon be possible with AI has come to be born. But that doesn't mean that deep learning is perfect. For example, it is known that there are many problems such as "inference is a black box", "unexpected answer by overfitting", and "large-scale network and long learning". Bayesian inference can provide learning and inference that is completely different from neural networks. Therefore, it might be possible to overcome the problems of neural networks. In this paper, we discuss the application of Bayesian inference to the estimation of parameters of the probability distribution. It also discusses the effects of prior probabilities and the relationship between Bayesian inference and maximum likelihood.
抄録全体を表示