This site is created for those who are interested in the modern state of advanced science in the field of artificial intelligence, neural networks and machine learning. On this site, students can find educational materials that will help them in studying topics related to neural networks.

We publish tutorials and blogs of various theoretical practical levels, ranging from the mathematical foundations of neural networks (backpropagation algorithm, recurrent neural networks, LSTM networks, deep learning algorithms, convolutional neural networks, etc.) to various examples neural networks application for solution of different practical problems (text processing, forecasting financial markets, image recognition, face recognition, etc.)

A **neuron** is an element that calculates the output signal (according to a certain rule) from a set of input signals. That is, the basic sequence of actions of one neuron is as follows:

- Receive signals from previous network elements.
- Combining input signals.
- Transmitting the output signal to the next elements of the neural network

Neurons can be connected in absolutely different ways, this is determined by the structure of a particular **neural network**. But the essence of the work of the neural network remains always the same. The set of signals arriving at the network input forms the output signal (or several output signals). That is, a neural network can be simplified in the form of a black box, which has inputs and outputs. And inside this box sits a huge number of neurons.

The source codes are available at: https://github.com/fgafarov/learn-neural-networks