We learn from our mistakes. And what about you?

Imagine that there were times when the thought process (you know, head is such an important part of a body we constantly use) was completely incomprehensible for the human mind. Before creating the artificial neural networks, people had to realize how their own mind works. The first 1,500 years the humanity was trying to study this issue as a philosophical matter. But when the mathematicians joined the research process, it became much more faster.

Nowadays every high school senior can draw an approximate scheme of a neural network. It is quite clear how it works. The next goal is to learn a neural network to do what people need (and what is beneficial). For example, to forecast cryptocurrency rate. If you summarize all the existing practices, you will see that it’s just the analysis or repeating of previous successful experiences. In general, it’s quite normal. As a result we have the forecasts with 75% accuracy. But how to improve the result? Our project can be described as “back to basics”. If the neural network resembles the human mind, so why don’t we learn it on that principle?

How do little children learn to walk? They stand up, try to take the first step but fall down. Then they try again and again. Finally they’ve got it! If you will hold them all the time or give a walker, that will automatically help to reach a positive result, but children will never walk on their own.

That is why we have decided not to make this mistake. We won’t invent walkers. We are going to rely on the evolutionary approach and try to find effective solutions to existing issues.