Machine Learning with WEKA Introduction and Preprocessing

1
Machine Learning with WEKAIntroduction and
Preprocessing
2
Outline

• How to preprocess data.
• How to add your own algorithm to Weka.
• How to use the test environment in Weka.

3
Resource

• UCI Machine Learning Repository
• ftp//ftp.ics.uci.edu/pub/machine-learning-databas
  es/
• Weka
• http//www.cs.waikato.ac.nz/ml/weka/
• http//www.cs.unb.ca/profs/hzhang/CS6735
• Tutorial.
• http//prdownloads.sourceforge.net/weka/weka.ppt

4
WEKA the software

• Machine learning/data mining software written in
  Java
• Main features
• Comprehensive set of data pre-processing tools,
  learning algorithms and evaluation methods
• Graphical user interfaces (incl. data
  visualization)
• Environment for comparing learning algorithms

5
Explorer pre-processing the data

• Data can be imported from a file in various
  formats ARFF, CSV, C4.5, binary
• Data can also be read from a URL or from an SQL
  database (using JDBC)
• Pre-processing tools in WEKA are called filters
• WEKA contains filters for
• Discretization, normalization, resampling,
  attribute selection, transforming and combining
  attributes,

6
WEKA only deals with flat files

• The data must be converted to ARFF format before
  applying any algorithm.
• The datasets name _at_relation
• The attribute information _at_attribute
• The data section begins with _at_data
• Data a list of instances with the attribute
  values being separated by commas.
• By default, the class is the last attribute in
  the ARFF file.

7
Numeric attribute and Missing Value

• _at_relation heart-disease-simplified
• _at_attribute age numeric
• _at_attribute sex female, male
• _at_attribute chest_pain_type typ_angina, asympt,
  non_anginal, atyp_angina
• _at_attribute cholesterol numeric
• _at_attribute exercise_induced_angina no, yes
• _at_attribute class present, not_present
• _at_data
• 63,?,typ_angina,233,no,not_present
• 67,male,asympt,286,yes,present
• 67,male,asympt,229,yes,present
• 38,female,non_anginal,?,no,not_present
• ...

The datasets name
The attribute information
Data
8
Numeric attribute and Missing Value

• _at_relation heart-disease-simplified
• _at_attribute age numeric
• _at_attribute sex female, male
• _at_attribute chest_pain_type typ_angina, asympt,
  non_anginal, atyp_angina
• _at_attribute cholesterol numeric
• _at_attribute exercise_induced_angina no, yes
• _at_attribute class present, not_present
• _at_data
• 63,?,typ_angina,233,no,not_present
• 67,male,asympt,286,yes,present
• 67,male,asympt,229,yes,present
• 38,female,non_anginal,?,no,not_present
• ...

numeric attribute
nominal attribute
Missing Value
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
Programming in Weka
Weka Test Environment
16
Add classifier into Weka

• Add your classifier class into \weka\gui\GenericOb
  jectEditor.props
• For example
• weka.classifiers.trees.J48,\
• weka.classifiers.YourOwnClassifier,\
• weka.classifiers.bayes.NaiveBayes,\

17
Performing experiments

• Experimenter makes it easy to compare the
  performance of different learning schemes
• For classification and regression problems
• Results can be written into file or database
• Evaluation options cross-validation, learning
  curve, hold-out
• Can also iterate over different parameter
  settings
• Significance-testing built in!

18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
END OF PART ONE