readme.txt

AdaBoost is an efficient tool in machine learning. It can combine a series of weak learners into a s

Description
=================
This source code is freely available for non-commercial use such as academic research and education. For other purpose, please contact me: [e-mail] 

AdaBoost is available at http://njustyw.googlepages.com/

AdaBoost is an efficient tool in machine learning. It can combine a series of weak learners into a strong learner. Besides pattern classification, it also can be applied into feature selection. This document explains the use of AdaBoost.


Installation
=================
On Unix/Linux systems, type 'make -f Makefile' to build the programs. Run them without arguments to see the usages of them.

On Windows systems, under Visual C++ environment, open 'AdaBoost.dsw' to build them.

The format of training and testing data file is:

<feature 1> <feature 2> ... <feature n> <label> 
<feature 1> <feature 2> ... <feature n> <label> 
.
.
.
where <feature 1>, <feature 2>, ..., <feature n> are the feature vector of each sample, <label> indicates the class label of each sample. 
A sample classification data included in this package is `data/train_data', you can obtain details.


Usage
=================
Training
(1): boost <max_iter> <trainfile> <modelfile>
        max_iter     -> maximum number of AdaBoost step (<int>)
        trainfile    -> the train file (<string>)
        modelfile    -> the model file of trained AdaBoost (<string>)

Testing
(2): boost <-test> <modelfile> <testfile>
        modelfile    -> the model file of trained AdaBoost (<string>)
        testfile     -> the test file (<string>)

Predicting
(3): boost <-predict> <modelfile> <predictfile> [outfile]
        modelfile    -> the model file of trained AdaBoost (<string>)
        predictfile  -> the predict file (<string>)
        outfile      -> the result file (<string>)


Examples
=================
> boost 100 data/train_data boost.model
Train a classifier with 100 iterations from the training data, and save trained model.

> boost -test boost.model data/test_data
Test the test file with trained model.

>boost -predict boost.model data/predict_data result.txt
Predict the predict file with trained model, and write result to a outfile.


=================
Copyright (c) 2006-2007, Great Yao.  Email:[e-mail]

http://njustyw.googlepages.com/