/tnn

🚧 Tiny Neural Network

Primary LanguageC++

Tiny Neural Network

Simple neural network implementation. Suitable for educational purposes.

Implementation allows you to add your own training methods, activation functions and network architecture.

Usage:

All work with the network will be done inside the file tnn.cpp.

🚧 Work in progress 🚧

Compile

I used g++ to compile. All work was done on Windows 10 Enterprise x64.

Windows

You can download g++ as part of MinGW here.

Unpack it so that the path to g++ looks like this

C:\\MinGW64\\bin\\g++.exe

or just edit the project settings in .vscode/launch.json and .vscode/tasks.json when you download it.

After all, open it via VS Code and Run!

GNU/Linux

🚧 Work in progress 🚧

Project tree

.
│   README.md             // This File
├───.vscode               // VS Code settings
│       launch.json
│       settings.json
│       tasks.json
├───bin
│       tnn.exe           // Output application file
├───doc                   // Docs folder
└───src
    │   tnn.cp           // Main file
    └───tnn              // Internal classes
            ActivationFuncSigmoid.hpp
            IActivationFunc.hpp
            IArch.hpp
            ILayer.hpp
            ILearnMethod.hpp
            Network.hpp
            Neuron.hpp
            SimpleLayer.hpp