I'm new to Weka and want to invoke classifiers and other Weka's capabilities from Matlab. Is there anyone who can help?

Mahnaz Behroozi @Mahnaz_Behroozi

05 May 2013 5 6K Report

I implemented the following piece of code.

clc;close all;clear all;

import weka.classifiers.Classifier

import weka.classifiers.bayes.BayesNet

import weka.classifiers.Evaluation;

v0 = java.lang.String('-I');

v01 = java.lang.String('200');

v02 = java.lang.String('-S');

v03 = java.lang.String('3');

v1 = java.lang.String('-t');

v2 = java.lang.String('C:\Program Files\Weka-3-6\data\iris.arff');

v3 = java.lang.String('-T');

v4 = java.lang.String('C:\Program Files\Weka-3-6\data\iris.arff');

v5 = java.lang.String('-W');

v6 =java.lang.String('weka.classifiers.trees.J48');

v06 = java.lang.String('--');

v7 = java.lang.String('-C');

v8 = java.lang.String('0.5');

prm = cat(1,v0,v01,v02,v03,v1,v2,v3,v4,v5,v6,v06,v7,v8);

output = Evaluation.evaluateModel(javaObject('weka.classifiers.meta.AdaBoostM1'),prm);

My main question is: How can I obtain outputs? (As you know the output with the means of "Evaluation" will be string). How can I get the confusion matrix from the output?

Looking forward to your replies.

Veronika Cheplygina

I'm not sure if this is directly what you need but perhaps this toolbox could help: http://www.mathworks.com/matlabcentral/fileexchange/21204-matlab-weka-interface

Mahnaz Behroozi

Thank you so much, dear Veronika. I tried to use the toolbox, but I faced this error:

"Arguments to IMPORT must either end with ".*"

or else specify a fully qualified class name:

"weka.core.converters.ArffLoader" fails this test."

It seems that there is a problem with importing the weka core! The new code I've tried to implement was aimed only to check the function "loadARFF" from the package you mentioned:

import weka.classifiers.*

import weka.core.converters.ArffLoader

fileName = java.lang.String('iris.arff');

data = loadARFF(fileName);

And the rest was the same. Any idea about the problem?

Have you tried import weka.core.converters.* instead? Actually this is already being done in loadARFF, so I don't think you have to call it separately. I'm not sure though, I have only used this toolbox a while ago and I don't have Weka installed anymore to test it out

Thanks a ton for your kind answers, Veronika. I'll try to find out the solution. Any other suggestions?

Saber Nankali

I have a question about running feature selection algorithms of weka in matlab

does Any body know how i can change the option of a feature selection algorithm of weka in matlab? for example i want to decrease the criteria of error to select more features.

Thanks

Is there any method or rule of thumb that can define the percentage of bearable missing data in one variable?

How can we find the optimum K in K-Nearest Neighbor?

How to optimize a combination of parameters using genetic algorithms?

Is there anyone who is expert in "cascade ensemble learning"? I need a powerful and preferably new survey on this area.

Timeout error!! Processor in the loop (PIL) using STM32F4 Matlab 2018a?

How to eliminate Steady state error from Reinforcement learning genetic algorithm code?

How to prevent ANFIS Model from negative prediction values in MATLAB?

How does an AC-DC power supply Flyback Converter Based USB Charger Model using Matlab/Simulink looks like?

Please. Could you help be in implementing Fractal Discrete Cosine Transform algorithm in Matlab?

D2D implementation in Matlab in Vienna LTE software?

Multi effect distillation modeling in EES?

What program can I use to caclulate dipole moment efficiently and without the need of excessive computation?

Does anyone ever solved this following Matlab error : "Error using imaqdevice/start (line 95) pointgrey: Failed to stop the camera acquisition." ?

How to build fouling and erosion simulation model for Aero Engine compressor and turbine rotor in Simulink / Matlab?