Interactive Visual Machine Learning Demos.

[EMNLP'20][Findings] Official Repository for the paper "Why and when should you pool? Analyzing Pooling in Recurrent Architectures."

Deep Neural Networks for music genre classification as a proxy for multiple analytical studies

Multilayer Perceptron GAN, and two Convolutional Neural Network GANs for MNIST and CIFAR.

Adaptive-saturated RNN: Remember more with less instability

This repository helps in understanding vanishing gradient problem with visualization

Code repository for my CSU master's research on dead ReLU's

Python codes for the experiment about LSTM Inefficiency in Long-Term Dependencies Regression Problems

This pdf shows some methods to fine tune the hyperparameters of a Neural Network to increase it's performance. It also puts light on some of the common problems related to Neural Networks along with their solutions.

This repository is based on the discussion about the Vanishing Gradient Problem. It explains some of the causes of this issue and provides solutions to help mitigate it.

Machine Learning Glossary

First public project written in Python i guess, anyway this is a repository for my class CS115.N12.KHCL

The vanishing gradient problem is a well-known issue in training recurrent neural networks (RNNs). It occurs when gradients (derivatives of the loss with respect to the network's parameters) become too small as they are backpropagated through the network during training.

Code for NeurIPS 2024 paper "Only Strict Saddles in the Energy Landscape of Predictive Coding Networks?"

Machine Learning Practical - Coursework 2: Analysing problems with the VGG deep neural network architectures (with 8 and 38 hidden layers) on the CIFAR100 dataset by monitoring gradient flow during training. And exploring solutions using batch normalization and residual connections.

Everything about Artificial Neural Network from Basic to Adavnced

I'll try to explain through the outcomes of Vanishing Gradient Problem

Machine Learning Practical - Coursework 2 Report: Analysing problems with the VGG deep neural network architectures (with 8 and 38 hidden layers) on the CIFAR100 dataset by monitoring gradient flow during training. And exploring solutions using batch normalization and residual connections.