Project: Classify Kaggle San Francisco Crime Description

Highlights:

• This is a multi-class text classification (sentence classification) problem.
• The goal of this project is to classify Kaggle San Francisco Crime Description into 39 classes.
• This model was built with CNN, RNN (LSTM and GRU) and Word Embeddings on Tensorflow.

Data: Kaggle San Francisco Crime

• Input: Descript

• Output: Category

• Examples:

  Descript	Category
  GRAND THEFT FROM LOCKED AUTO	LARCENY/THEFT
  POSSESSION OF NARCOTICS PARAPHERNALIA	DRUG/NARCOTIC
  AIDED CASE, MENTAL DISTURBED	NON-CRIMINAL
  AGGRAVATED ASSAULT WITH BODILY FORCE	ASSAULT
  ATTEMPTED ROBBERY ON THE STREET WITH A GUN	ROBBERY

Train:

• Command: python3 train.py train_data.file train_parameters.json
• Example: python3 train.py ./data/train.csv.zip ./training_config.json

Predict:

• Command: python3 predict.py ./trained_results_dir/ new_data.csv
• Example: python3 predict.py ./trained_results_1478563595/ ./data/small_samples.csv

Server:

• Command: python3 server.py ./trained_results_dir
• Open http://127.0.0.1:5000/ (Press CTRL+C to quit)

References:

• Implement a cnn for text classification in tensorflow
• See .staging/notes/REFERENCES.md

Docker

Docker Compose Example

======================

This is Docker Compose example. You can boot a container based on this repository image with Jupyter port 8888, its password 'foobar' and sharing files in notebook directory with the container by just running

docker-compose up

To stop it, please press 'Ctrl-C'.