https://github.com/iglee/perceptron-text-classifier
Sentiment Classification exercise with perceptron, feed-forward multilayer net, LSTM RNN, and RCNN!
https://github.com/iglee/perceptron-text-classifier
bigram feedforward-neural-network lstm perceptron pytorch rcnn rnn unigram
Last synced: 3 months ago
JSON representation
Sentiment Classification exercise with perceptron, feed-forward multilayer net, LSTM RNN, and RCNN!
- Host: GitHub
- URL: https://github.com/iglee/perceptron-text-classifier
- Owner: iglee
- Created: 2020-02-17T08:01:07.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2020-03-02T19:02:55.000Z (over 6 years ago)
- Last Synced: 2025-05-14T18:53:56.685Z (about 1 year ago)
- Topics: bigram, feedforward-neural-network, lstm, perceptron, pytorch, rcnn, rnn, unigram
- Language: Python
- Homepage:
- Size: 575 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
## Perceptron Sentiment Classification of Text
### 1. Single Perceptron Sentiment Classification
- Pre-processing: lowercasing, frequency of word count to limit vocabulary, NLTK Snowball stemming
- Unigram Feature
- Unigram and Bigram Features
### 2. Multilayer Perceptron Sentiment Classification
- two layer with tanh activation layer, with sigmoid activation for final output (used BCE loss)
### 3. Recurrent Neural Networks
- Bi-directional LSTM with "max pooling" activation to a linear fully connected layer. Then, this is fed to final Sigmoid activation.
### 4. RCNN
- Bi-directional LSTM with convolution and max pooling.