Switch branches/tags
Nothing to show
Find file History
Latest commit a3f5aa9 Jan 12, 2015


3: Text Classification


  • Python 2.7.x
  • Module: See requirements.txt


Python Modules

pip install -r requirements.txt

Preprocessing Data

Download data from the official competition page and rename it data.tar.gz.

tar xvzf data.tar.gz

Run solvers

Run SVM and Gradient Boosting Classifier. This process requires more than 10 hours.

python solvers/svm.py data/bag-of-words/train_tfidf.svmlight data/bag-of-words/test_tfidf.svmlight -o pred_svm -v
python solvers/gbc.py data/bag-of-words/train_tfidf_lsi300.svmlight data/bag-of-words/test_tfidf_lsi300.svmlight -o pred_gbc -v

Blend results

python scripts/blend.py pred_svm pred_gbc -o pred

Submit pred and be happy!