The document provides an agenda for a workshop on data mining using the WEKA tool. The agenda covers an introduction to data mining and WEKA, as well as demonstrations of WEKA's capabilities for classification analysis, predictive analysis, cluster analysis, and hands-on exercises. An overview of WEKA is also given, explaining that it is open-source software for data pre-processing, machine learning, and evaluating algorithms.

1. Data Mining using WEKA tool
Dr. S KANNIMUTHU,
Professor,
Head-Center of Excellence in Algorithms,
Department of CSE,
Karpagam College of Engineering,
Coimbatore.
1 10/30/2022
WORKSHOP ON DATA MINING
USING WEKA AND PYTHON

2. AGENDA
 Introduction to Data Mining and Tools
 Introduction to WEKA Tool
 WEKA-Explorer
 WEKA-Experimenter
 WEKA-Knowledgeflow
 Classification Analysis
 Predictive Analysis
 Cluster Analysis
 Hands-on
10/30/2022
2

3. Data Mining
 Data mining is the exploration and analysis of
large quantities of data in order to discover
valid, novel, potentially useful, and ultimately
understandable patterns in data.
Valid: The patterns hold in general.
Novel: We did not know the pattern beforehand.
Useful: We can devise actions from the patterns.
Understandable: We can interpret and
comprehend the patterns.
10/30/2022
3

4. What are data mining tools?
 Data mining tools are software components and
theories that allow users to extract information
from data.
 The tools provide individuals and companies with
the ability to gather large amounts of data and
use it to make determinations about a particular
user or groups of users.
 Some of the most common uses of data mining
tools are in the fields of marketing, healthcare
system, fraud protections and surveillance.
10/30/2022
4

5. WEKA: the software
5
 Machine learning/data mining software written in
Java (distributed under the GNU Public License)
 Used for research, education, and applications
 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
10/30/2022

6. WEKA only deals with “flat” files
6
@relation heart-disease-simplified
@attribute age numeric
@attribute sex { female, male}
@attribute chest_pain_type { typ_angina, asympt, non_anginal, atyp_angina}
@attribute cholesterol numeric
@attribute exercise_induced_angina { no, yes}
@attribute class { present, not_present}
@data
63,male,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
...
10/30/2022

7. WEKA only deals with “flat” files
7
@relation heart-disease-simplified
@attribute age numeric
@attribute sex { female, male}
@attribute chest_pain_type { typ_angina, asympt, non_anginal, atyp_angina}
@attribute cholesterol numeric
@attribute exercise_induced_angina { no, yes}
@attribute class { present, not_present}
@data
63,male,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
...
10/30/2022

8. 8
java weka.gui.GUIChooser
10/30/2022

9. 9
java -jar weka.jar
10/30/2022

10. Explorer: pre-processing the data
10
 Data can be imported from a file in various
formats: ARFF, CSV, C4.5, binary
 Pre-processing tools in WEKA are called “filters”
 WEKA contains filters for:
 Discretization, normalization, resampling, attribute
selection, transforming and combining attributes, …
10/30/2022

11. 11 10/30/2022

12. 12 10/30/2022

13. 10/30/2022
13

14. 14 10/30/2022

15. 15 10/30/2022

16. 16 10/30/2022

17. 10/30/2022
17

18. 18 10/30/2022

19. Explorer: building “classifiers”
19
 Classifiers in WEKA are models for predicting
nominal or numeric quantities
 Implemented learning schemes include:
 Decision trees and lists, instance-based classifiers,
support vector machines, multi-layer perceptrons,
logistic regression, Bayes’ nets, …
 “Meta”-classifiers include:
 Bagging, boosting, stacking, error-correcting output
codes, locally weighted learning, …
10/30/2022

20. 10/30/2022
20

21. 10/30/2022
21

22. 10/30/2022
22

23. 10/30/2022
23

24. 10/30/2022
24

25. 10/30/2022
25

26. 10/30/2022
26

27. 10/30/2022
27

28. 10/30/2022
28

29. 10/30/2022
29

30. 10/30/2022
30

31. 10/30/2022
31

32. 10/30/2022
32

33. 10/30/2022
33

34. 10/30/2022
34

35. 10/30/2022
35

36. 10/30/2022
36

37. 10/30/2022
37

38. 10/30/2022
38

39. 10/30/2022
39

40. 40 10/30/2022

41. 10/30/2022
41

42. 10/30/2022
42

43. 10/30/2022
43

44. 10/30/2022
44

45. 10/30/2022
45

46. 10/30/2022
46

47. 10/30/2022
47

48. 10/30/2022
48

49. 10/30/2022
49

50. 10/30/2022
50

51. 10/30/2022
51

52. 10/30/2022
52

53. 10/30/2022
53

54. 10/30/2022
54

55. 10/30/2022
55

56. 10/30/2022
56
QuickTime™ and a TIFF (LZW) decompressor are needed to see this picture.

57. 10/30/2022
57
QuickTime™ and a TIFF (LZW) decompressor are needed to see this picture.

58. 10/30/2022
58

59. 10/30/2022
59

60. 60 10/30/2022

61. 10/30/2022
61

62. 10/30/2022
62
Quick T
ime™an
d aT
IFF(L
ZW)de
c o
mpre
s sorar
e n
ee
de
dtos e
e t
his p
ic t
ure
.

63. 10/30/2022
63

64. 10/30/2022
64

65. 10/30/2022
65

66. 10/30/2022
66

67. 10/30/2022
67

68. 10/30/2022
68

69. 10/30/2022
69

70. 10/30/2022
70

71. 10/30/2022
71

72. 10/30/2022
72

73. 10/30/2022
73

74. 74 10/30/2022

75. 75 10/30/2022

76. 10/30/2022
76

77. Explorer: clustering data
77
 WEKA contains “clusterers” for finding groups of
similar instances in a dataset
 Some implemented schemes are:
 k-Means, EM, Cobweb, X-means, FarthestFirst
 Clusters can be visualized and compared to “true”
clusters (if given)
10/30/2022

78. 10/30/2022
78

79. 10/30/2022
79

80. 10/30/2022
80

81. 10/30/2022
81

82. 10/30/2022
82

83. 10/30/2022
83

84. 10/30/2022
84

85. 10/30/2022
85

86. 10/30/2022
86

87. Explorer: finding associations
87
 WEKA contains the Apriori algorithm (among
others) for learning association rules
 Works only with discrete data
 Can identify statistical dependencies between
groups of attributes:
 milk, butter  bread, eggs (with confidence 0.9 and
support 2000)
 Apriori can compute all rules that have a given
minimum support and exceed a given confidence
10/30/2022

88. 10/30/2022
88

89. 10/30/2022
89

90. 10/30/2022
90

91. Explorer: attribute selection
91
 Panel that can be used to investigate which
(subsets of) attributes are the most predictive
ones
 Attribute selection methods contain two parts:
 A search method: best-first, forward selection,
random, exhaustive, genetic algorithm, ranking
 An evaluation method: correlation-based, wrapper,
information gain, chi-squared, …
 Very flexible: WEKA allows (almost) arbitrary
combinations of these two
10/30/2022

92. 10/30/2022
92

93. 10/30/2022
93

94. 10/30/2022
94

95. 10/30/2022
95

96. 10/30/2022
96

97. 10/30/2022
97

98. Explorer: data visualization
98
 Visualization very useful in practice: e.g. helps to
determine difficulty of the learning problem
 WEKA can visualize single attributes (1-d) and
pairs of attributes (2-d)
 Color-coded class values
 “Jitter” option to deal with nominal attributes (and
to detect “hidden” data points)
 “Zoom-in” function
10/30/2022

99. 10/30/2022
99

100. 10/30/2022
100

101. 10/30/2022
101

102. Performing experiments
102
 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
10/30/2022

103. 10/30/2022
103

104. 10/30/2022
104

105. 10/30/2022
105

106. 106 10/30/2022

107. 10/30/2022
107

108. 10/30/2022
108

109. 10/30/2022
109

110. 10/30/2022
110

111. The Knowledge Flow GUI
111
 Java-Beans-based interface for setting up and
running machine learning experiments
 Data sources, classifiers, etc. are beans and can
be connected graphically
 Data “flows” through components: e.g.,
“data source” -> “filter” -> “classifier” ->
“evaluator”
 Layouts can be saved and loaded again later
10/30/2022

112. 112 10/30/2022

113. 113 10/30/2022

114. Conclusion:
114
 WEKA is available at
http://www.cs.waikato.ac.nz/ml/weka
 Also has a list of projects based on WEKA
 (probably incomplete list of) WEKA contributors:
Abdelaziz Mahoui, Alexander K. Seewald, Ashraf M. Kibriya,
Bernhard Pfahringer, Brent Martin, Peter Flach, Eibe Frank, Gabi
Schmidberger, Ian H. Witten, J. Lindgren, Janice Boughton,
Jason Wells, Len Trigg, Lucio de Souza Coelho, Malcolm Ware,
Mark Hall, Remco Bouckaert, Richard Kirkby, Shane Butler,
Shane Legg, Stuart Inglis, Sylvain Roy, Tony Voyle, Xin Xu, Yong
Wang, Zhihai Wang
10/30/2022

115. Reference
 Ian H.Witten & Eibe Frank, “Data mining-Practical
Machine Learning Tools and Techniques”,Elsevier
publication, Second Edition
 Weka Manual
10/30/2022
115

116. Contact
 WhatsApp: 8838941811
 Email: drkannimuthu@gmail.com
 Linkedin: https://www.linkedin.com/in/kannimuthu-
subramanian-292924a/
 Google Scholar:
https://scholar.google.co.in/citations?user=eSdX5
S0AAAAJ&hl=en
 Publon:
https://publons.com/researcher/1686169/kannimu
thu-subramanian/
 Blog: www.skannimuthu.blogspot.in
 Github: www.github.com/kannimuthu 10/30/2022
116

117. THANK YOU
10/30/2022
117