Skip to content
forked from lrvine/Bayesian

Naive Bayes classifier and Bayesian network classifier C++ implementation

License

Notifications You must be signed in to change notification settings

baosince/Bayesian

 
 

Repository files navigation

Bayesian

Naive Bayes classifier and Bayesian network classifier C++ implementation

//TODO refactor the code to smaller functions and rename variables to be more meaningful

Reference :

Naive Bayes Classifier

Bayesian Network

Usage:

./bayesian [training data file] [input test file] [configuration file] [method]

method :

0 : naive bayesian
1 : bayesian network

Example:

./bayesian data.txt test.txt cfg.txt 0

Training & Test data are in CSV format:

[ attribute 1 of data 1 ],[ attribute 2 of data 1 ], ........ ,[ result of data 1 ]
[ attribute 1 of data 2 ],[ attribute 2 of data 2 ], ........ ,[ result of data 2 ]
[ attribute 1 of data 3 ],[ attribute 2 of data 3 ], ........ ,[ result of data 3 ]

Configuration file format:

[ number of training instance ]  [ number of test instance ]  [ number of attributes ]

[ attribute 1 is discrete(0) or continunous(1) ]  [ attribute 2 is discrete(0) or continunous(1) ]  ......

[ number of types for attributes 1 ]  [ number of types for attributes 2 ]  .......[ number of classes for prediction result ]  

* For continuous data, the "number of types for attributes" won't matter.
* Bayesian Network DO NOT support continuous data for now.

Data set is acquired from UCI Machine Learning Repository

About

Naive Bayes classifier and Bayesian network classifier C++ implementation

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 98.3%
  • Makefile 1.7%