File Exchange

image thumbnail

Ensemble Toolbox

version 1.0 (1.42 MB) by

This toolbox provides some combination methods to fuse an ensemble of classifiers

32 Downloads

Updated

View License

Please cite at least one of the following papers if you like to use this toolbox:
+ M.A. Bagheri, Gh. Montazer, and E. Kabir, “A Subspace Approach to Error-Correcting Output Coding”,
Pattern Recognition Letters, vol. 34, pp. 176–184, 2013
+ M.A. Bagheri, Q. Gao and S. Escalera, “Rough Set Subspace Error Correcting Output Codes”,
in Proc. IEEE International Conf on Data Mining, Brussels, Belgium, 2012

For better understanding of the toolbox, reading the following references is strongly recommended
[1] R. Polikar, "Ensemble based systems in decision making," IEEE Circuits and Systems Magazine, vol. 6, pp. 21-45, 2006.
[2] L. I. Kuncheva, Combining Pattern Classifiers: Methods and Algorithms. New York, NY: Wiley, 2004.

%%%% How to use
%% You should run the Main_Ensemble.m file

%% Dataset
*** You can use any of the included datasets (taken from UCI ) or your own dataset;
*** To use your dataset, you should put the dataset in the Datasets folder
*** The data should be stored in one text file. You can save data in "Excel > Save as Text (Tab Delimited)"
*** The dataset should be in a Matrix format;
*** number of rows is equal to the number of samples
*** Number of columns is equal to the number of features + 1
*** The last column is the Target (Label) column of samples

Comments and Ratings (18)

Omokhadija

Thank you very much for this tools.
the Dempster shafer fusion is not normalized here with the conflict factor? Can you please explain why??

Farrikh zami

i think this toolbox didn't have the bagging method (bootstrap the dataset).

essam

essam (view profile)

can you tell me how to ensemble my custom neural network and custom KNN together using DT as top level/stacked model learner?

Farrikh zami

@surendra singh, in LoadData.m try this RawData=load(strcat('Datasets/',dataset));
you just change \ become /.
it works for me

Jialin Yu

This is a very useful work. But in CombineCLFs file, there are two 'case {9}', one is for BKS and the other is for DS. In BKS, there is an undefined variable Train_CM_ind.

Michelle

Mr. Sumon

myknnclassify.m file is missing please check!!!!

Error using load
Unable to read file Datasets\Glass.txt: No such file or directory.

Error in LoadData (line 5)
RawData=load(strcat('Datasets\',dataset));

Error in Main_Ensemble (line 42)
[Patterns,Targets,V_Targets]=LoadData(dataset,scale);

MATLAB VERSION:R2012a
#Pls. Ignore previous comment..

Error using load
Unable to read file Datasets\e:\Glass.txt.txt: Invalid argument.

Error in LoadData (line 5)
RawData=load(strcat('Datasets\',dataset));

Error in Main_Ensemble (line 42)
[Patterns,Targets,V_Targets]=LoadData(dataset,scale);

MATLAB VERSION:R2012a

DSP

DSP (view profile)

Can I replace other neural network classifier (e.g RBFNN, PNN etc ) as base classifier instead of KNN and SVM? Thanks

vaishali

Ensemble toolbox is very useful for my research.I want to use KNN classifier along with other classifiers.
Not finding any files related in KNN toolbox. Main ensemble is not running with KNN(2).
Can you guide me in this regard.

Anita, could you please let me know what's the error?
Also, please note that some combination techniques are not inherently suitable for KNN (classifier type=2).
Finally, you may send me the file you are running or the whole package...

Anita

Anita (view profile)

this package is really outstanding... when i tried with ClassifierTypes = [2 2 2], it is showing some error in accuracy... please guide me to use this toolbox for my dataset.

Hi. In this package, I did not include different features selection techniques. I will submit another package in the next month that includes about 4-5 feature selection methods. Meanwhile, some of them are already available in MatlabFileExchange (not by me! )

alex hsu

It's outstanding work,but how can i use other feature selection methods??(only all&random)~thanks

alex hsu

alex hsu

MATLAB Release
MATLAB 7.6 (R2008a)

Download apps, toolboxes, and other File Exchange content using Add-On Explorer in MATLAB.

» Watch video