Document Classification ML

• BBC news dataset has been used to classify models
  • Consists of 2225 documents from the BBC news website corresponding to stories in five topical areas from 2004-2005.
  • Natural Classes: 5 (business, entertainment, politics, sport, tech)
  • http://mlg.ucd.ie/datasets/bbc.html
  • D. Greene and P. Cunningham. "Practical Solutions to the Problem of Diagonal Dominance in Kernel Document Clustering", Proc. ICML 2006.
• Models trained: Naive Bayes, KNN, SVM and Random Forest
• GridSearch with cross validation = 10

Training the model

• place the data files in the data folder
• Run main.py to train the model
• The models will save to save_model folder
• results and metrics will be displayed on console

Predicting using a saved model

• run test_predict.py by proving the full path to the saved model file
• results will be displayed on console