Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
250 views12 pages

Weka 3.6 Data Mining Guide

Weka is a machine learning tool developed by the University of Waikato that provides algorithms for data preprocessing, classification, regression, clustering, association rule mining and feature selection. It has a graphical user interface called Explorer for loading and preprocessing data files and evaluating models, as well as command line interfaces. Weka supports common data formats like ARFF and CSV and can preprocess data using filters before applying classification and other machine learning algorithms. Example commands are provided for loading data, applying classifiers like J48 decision trees and Naive Bayes, and using filters.

Uploaded by

Mehdi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
250 views12 pages

Weka 3.6 Data Mining Guide

Weka is a machine learning tool developed by the University of Waikato that provides algorithms for data preprocessing, classification, regression, clustering, association rule mining and feature selection. It has a graphical user interface called Explorer for loading and preprocessing data files and evaluating models, as well as command line interfaces. Weka supports common data formats like ARFF and CSV and can preprocess data using filters before applying classification and other machine learning algorithms. Example commands are provided for loading data, applying classifiers like J48 decision trees and Naive Bayes, and using filters.

Uploaded by

Mehdi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

Weka 3.

6 Tutorial
(Waikato Environment for Knowledge Analysis)
WEKA
• It’s a data mining/machine learning tool developed
by University of Waikato.

• Main Features:
– 49 data preprocessing tools
– 76 classification/regression algorithms
– 8 clustering algorithms
– 3 algorithms for finding association rules
– 15 attribute/subset evaluators + 10 search algorithms for
feature selection
Starting WEKA
• 4 Options
– Explorer
– Experimenter
– Knowledge Flow
– Simple CLI
Weka Simple CLI

Commands here
Preprocessing Data
• Data can be imported from a file in various
formats: ARFF, CSV etc.
• Data can also be read from a URL or from an
SQL database (using JDBC)
• Pre-processing tools in WEKA are called
“filters”
CSV (Comma Separated) File
Roll,Name,Percentage,Passed
1,ABC,72,y
2,abc,30.8,n
Result.csv
3,xyz,44.3,n
4,XYZ,52.3,y

Command for converting csv to arff:

java weka.core.converters.CSVLoader Result.csv > Result.arff


ARFF File Description of dataset

@relation Result
Numeric

@attribute Roll numeric


@attribute Name {ABC,abc,xyz,XYZ}
Nominal
@attribute Percentage numeric
@attribute Passed {y,n}

@data
1,ABC,72,y
2,abc,30.8,n weka.core.Instance
3,xyz,44.3,n
4,XYZ,52.3,y
Weka.Classifiers
• A simple example

java weka.classifiers.trees.J48 -t data/weather.arff

Classifier: Decision tree


Specifies training file

Classifier: NaiveBayes

java weka.classifiers.bayes.NaiveBayes -t data/weather.arff


Weka.Classifiers
• Classifier’s option:
– weka.classifiers.trees.J48
– weka.classifiers.bayes.NaiveBayes
– weka.classifiers.functions.Logistic
– weka.classifiers.functions.SMO
Weka.Classifiers
• Other options:
– -t : Specifies training file
– -T: Specifies test file
– -x: Number of cross validation
– -c: set the class index
– -d: Save the model
– -i: Detailed performance description
– -p #: Prediction and one attribute value
Few more complex examples
java -Xmx1024m weka.classifiers.trees.J48 -t data.arff –i –x 3

java -Xmx1024m weka.classifiers.trees.J48 -t data.arff –i –T


testData.arff –p 1

java -Xmx1024m weka.classifiers.trees.J48 -t data.arff –i –T


testData.arff -d J48-data.model >&! J48-data.out &

java -Xmx1024m weka.classifiers.meta.FilteredClassifier -t data.arff –i


–T testData.arff –i –c last –F
weka.filters.unsupervised.attribute.StringToWordVector –W
weka.classifier.functions.SMO > Output

java -Xmx1024m –cp weka.jar:LibSVM.jar


weka.classifiers.meta.FilteredClassifier -t data.arff –i –T
testData.arff –i –c last –F
weka.filters.unsupervised.attribute.StringToWordVector –W
weka.classifier.functions.LibSVM > Output
Weka Filters
• Used to transform input data
– Removing or adding attributes
– Resampling the dataset
– Removing examples
– ...

You might also like