Heterogeneous automatic differentiation ("backpropagation") in Haskell

Фреймворк для построения нейронных сетей, комитетов, создания агентов с параллельными вычислениями.

This repository contains all my theroy reports, written assignments and programming code that I wrote/referrd for the DL course at IIT,Madras taught my advisor Prof.Mitesh Khapra.

Code for SEDONA: Search for Decoupled Neural Networks toward Greedy Block-wise Learning (ICLR 2021)

For Azimuth ACT course

基于并行BP神经网络的人脸识别系统（终端）

Lecture notes on backpropagation

Using only numpy in Python, a neural network with a forward and backward method is used to classify given points (x1, x2) to a color of red or blue.

I have designed these neural networks to revise the mathematics involved in their training, I have derived by hand all of the backprop and learning equations. These neural networks may not be the most efficient but, efficiency was not the aim here. The aim here was understanding. The various networks contained include a Rosenblatt Perceptron, a …

A Simple Neural Network Engine in c++ which implements back-propagation algorithm it contains lots of flaws and it is intended to used for fun only

Backpropagation in Neural Network (NN) with Python

perceptron, backprop, RBF, SOM, hopfield nets, autoencoders (no external ML libs)

Neural Net Library w/ OpenCL Backend

The classic Kaggle Titanic data science challenge

Small-scale implementation of an automatic differentiation engine, inspired by PyTorch's autograd system

Numpy neural network classifying japanese characters. Weights animations along the way.