TamingAI is an educational project about the technology of neural networks. Consisting of two chapters, it covers the building blocks of a typical machine learning algorithm and explains how to apply them to a real project.

We also include a simple neural network that is made to classify handwritten digits using the MNIST dataset.

The algorithm itself is written from the ground up in C++. Therefore, we did not use any external libraries to build neural networks such as OpenNN, Dlib etc.

TamingAI makes use of a variety of open source projects including:

TamingAI is ideal for people who don’t only want to learn about how to use AI, but also how it works under the hood.

Learn more and contribute on GitHub: TamingAI, NeuralNetwork.