Skip to content

Latest commit

 

History

History
16 lines (12 loc) · 671 Bytes

README.md

File metadata and controls

16 lines (12 loc) · 671 Bytes

nb-tan-classfiers

A program that implements both Naive Bayes and TAN (tree-augmented naive Bayes)

This python program implements the Naive Bayes and TAN classifiers intended for binary classification.

  • Uses Prim algorithm to construct the Maximal Spanning Tree in TAN.
  • Laplace estimates is used to calculate all probabilities.
  • The classifier can be tested on data sets used to predict lymphatic cancer included in the repo.
  • The file format expected for the training and testing data sets is ARFF : http://www.cs.waikato.ac.nz/ml/weka/arff.html

Usage

bayes <train-set-file> <test-set-file> <n|t>

where n indicates Naive Bayes, t indicates TAN.