Simplified implementations of some interesting works related to deep learning.
Comparison of various SGD algorithms on logistic regression and MLP. The relation of these algorithms is shown in the following figure, please refer to sgd-comparison for the details.
Use neural network to approximate functions. The approximated functions are shown in the following figures, please refer to function-approximation for the details.
Generative Adversarial Networks
A simple demonstration of Generative Adversarial Networks (GAN), maybe problematic.https://arxiv.org/abs/1406.2661), we also use GAN to generate Gaussian distribution which shown in the left figure. Then we try to generate digits based on MNIST dataset, however, we encouter "the Helvetica scenario" in which G collapses too many values of z to the same value of x. Nevertheless, it is a simple demonstration, please see the [details](https://github.com/SwordYork/simplified-deeplearning/tree/master/GAN).