A neural network, which is said to have resulted in a revolutionary evolution in the development of AI. It is said that it made learning closer to humans possible by imitating the mechanism of human brain cell "neuron", and the activity of AlphaGo which is the software of go is remembered in memory.

At IBM 's Web media infinity, a new technology to further evolve the neural network was introduced. Its name is dynamic Boltzmann machine.

This is mainly related to memory. In the first place, the human brain seems to be more strongly memorized when two neurons are stimulated at the same time. For example, when you see a cat and listen to the word cat! It seems that deep learning and memory is urged. There is a sense of satisfaction.

The dynamic Boltzmann machine is focused on the fact that the " time difference " is related to the strength of the binding of this cell, and it is an AI technology that enables prediction of data that changes with time and abnormality detection.

Specifically, it seems that it is expected to be applied to monitoring the heart rate, blood pressure, body temperature, etc. of a person to manage the physical condition, predicting a crash or a sharp rise from stock price movements.

There are many technical terms and it is somewhat difficult contents, but if you are interested in dynamic Boltzmann machine and new possibilities of the neural network.