N-Layer Neural Network Framework
C++Machine LearningNeural NetworksBackpropagationDeep LearningGradient Descent
Project Overview
A flexible N-layer feed-forward neural network framework implemented entirely from scratch in C++. Supports configurable network architectures with multiple hidden layers, various activation functions (sigmoid, tanh, linear), and backpropagation training with gradient descent. Features external configuration files, weight saving/loading, truth table management, keep-alive training monitoring, and image classification for handwritten digit recognition with adjustable learning rates and error thresholds.
Screenshots
