https://github.com/yashk2000/bollerenverse

A few things I learned while learning about machine learning.
https://github.com/yashk2000/bollerenverse

acer deep-learning lstm neural-networks python3 pytorch reinforcement-learning rnn-pytorch stable-baselines

Last synced: 18 days ago
JSON representation

A few things I learned while learning about machine learning.

• Host: GitHub
• URL: https://github.com/yashk2000/bollerenverse
• Owner: yashk2000
• Created: 2020-02-14T17:06:09.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2021-04-26T11:43:14.000Z (over 3 years ago)
• Last Synced: 2024-10-28T16:55:04.287Z (2 months ago)
• Topics: acer, deep-learning, lstm, neural-networks, python3, pytorch, reinforcement-learning, rnn-pytorch, stable-baselines
• Language: Jupyter Notebook
• Homepage:
• Size: 469 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# BolleRenverse

**1) [SimpleRNN](https://github.com/yashk2000/BolleRenverse/blob/master/SimpleRNN.ipynb)**

This is a simple RNN, a recurrent neural network whch does time series prediction. Given some set of input data, it is able to generate a prediction for the next time step.

**2) [Sentiment Analysis](https://github.com/yashk2000/BolleRenverse/blob/master/SentimentAnalysis.ipynb)**

This RNN is used to detect the sentiment in a given moview reivew. Given a movie review as an input, it classifes the review as *positive* or *negative*. It used LSTM, Long Short Term Memory to perform the analysis. 

**3) [MNIST GAN](https://github.com/yashk2000/BolleRenverse/blob/master/MNISTGan.ipynb)**

This is a GAN, Generative Adverserial Network used to generate hand written digits. It consists of a a discriminator neural network trained on the MNIST Handwritten digits dataset and a generator neural network which generates new images. If the image is classified as a hand written digit by the discirminator, then we have a digit successfully generated by the GAN. 

**4) [Character Generating RNN](https://github.com/yashk2000/BolleRenverse/blob/master/characterGeneratingRNN/charRNN.ipynb)**

This is a RNN which generates sentences character by character after being trained on a input text which in this case is *Percy Jackson and the Last Olympian*. It requires an input prime, which can be a single string or a collection of strings and generates sentences using those word/s.

**5) [Diner Dash using reinforcement learning](https://github.com/yashk2000/BolleRenverse/blob/master/DinerDash.ipynb)**

[Diner dash](https://github.com/AdaCompNUS/diner-dash-simulator) is a game in which the player has to run a restaurant in a way to keep maximum customers happy. This is an attempt to play the game using reinforcement leanring using Stable Basline functions. This solution uses [ACER(Actor-Critic with Experience Replay)](https://arxiv.org/abs/1611.01224).