Professional Documents
Culture Documents
1|Page
Assignment: 1
Introduction
Gathered data sets are include valuable information and knowledge which is often hidden. Processing the huge data
and retrieving meaningful information from it is a difficult task. The aim of our work is to investigate the performance
of different classification methods using WEKA for different three dataset obtained from UCI data archive.
WEKA is an open source software which consists of a collection of machine learning algorithms for data mining tasks.
This assignment is to investigate the performance of different classification or clustering methods for a set of large
data set.
ii.
iii.
Multilayer Perceptron: Multilayer Perceptron is a nonlinear classifier based on the Perceptron. A Multilayer
Perceptron (MLP) is a back propagation neural network with one or more layers between input and
output layer.
A Support Vector Machine (SVM): SVM is a discriminative classifier formally defined by a separating hyper
plane. In other words, given labeled training data (supervised learning), the algorithm outputs an optimal
hyper plane which categorizes new examples.
J48: The J48 algorithm is WEKAs implementation of the C4.5 decision tree learner. The algorithm uses a
greedy technique to induce decision trees for classification and uses reduced-error pruning.
2|Page
Assignment: 1
iv.
IBk: IBk is a k-nearest-neighbor classifier that uses the same distance metric. k-NN is a type of instance
based learning or lazy learning where the function is only approximated locally and all computation is
deferred until classification. In this algorithm an object is classified by a majority vote of its neighbors.
v.
Naive Bayesian: Naive Bayesian classifier is developed on bayes conditional probability rule used for
performing classification tasks, assuming attributes as statistically independent; the word Naive means
strong. All attributes of the data set are considered as independent and strong of each other.
Steps to apply classification techniques on data set and get result in Weka:
Classification
Multilayer
Perceptron
Support Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
69.56
99.7299
0.2701
0.996
0.0014
0.0186
0.5218
5.0233
14.23
97.5617
2.4383
0.9641
0.0098
0.0988
3.5721
26.7298
0.03
97.0525
2.9475
0.9568
0.0153
0.0951
5.6151
25.7324
98.3796
1.6204
0.9761
0.0859
0.1466
31.474
39.6775
90.3241
9.6759
0.8567
0.0765
0.1767
28.0234
47.8152
Classification
Multilayer
Perceptron
Support Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
69.28
97.4353
2.5647
0.962
0.006
0.0514
2.1843
13.9063
8.62
97.4353
2.5647
0.962
0.006
0.0514
2.1843
13.9063
0.14
96.4821
3.5179
0.9483
0.0186
0.1055
6.7947
28.5491
97.5261
2.4739
0.9636
0.0854
0.1512
31.2706
40.9314
0.03
90.6718
9.3282
0.8618
0.077
0.1766
28.185
47.7877
3|Page
Assignment: 1
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Cross validation 10
Naive
Bayesian
66% split
J48
k-nearest
neighbor
Cross validation 10
Naive
Bayesian
102
100
98
96
94
92
90
88
86
84
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Cross validation 10
Naive
Bayesian
66% split
66% split
Capa Statistics
0.1
0.08
0.06
0.04
0.02
0
Multilayer Support
Perceptron Vector
Machine
J48
Cross validation 10
4|Page
k-nearest
neighbor
Naive
Bayesian
66% split
1.05
1
0.95
0.9
0.85
0.8
0.75
Multilayer Support
Perceptron Vector
Machine
J48
Cross validation 10
k-nearest
neighbor
Naive
Bayesian
66% split
Assignment: 1
60
50
40
30
20
10
0
Multilayer Support
Perceptron Vector
Machine
RMs
J48
RAE
k-nearest
neighbor
Multilayer Support
Perceptron Vector
Machine
Naive
Bayesian
RMs
RRSE
J48
RAE
k-nearest
neighbor
Naive
Bayesian
RRSE
Above two graphs are showing comparison of different error parameters, considerably the multilayer
perception classifier showing good results, it means lower error rate. Except others, but k-nearest and nave
Bayesian are showing high amount of error in determining the five classes.
10-fold cross-validation
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Time
taken
seconds
Correctly
Classified
Instances
Multilayer
Perceptron
Incorrectly
Classified
Instances
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Root relative
squared
error
Naive
Bayesian
Correctly
Classified
Instances
Multilayer
Perceptron
Incorrectly
Classified
Instances
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Root relative
squared
error
Naive
Bayesian
The above two graphs are showing the compared performance matrices of classifiers in percentage. The close look
of these graphs are showing no significant changes between the parameters. The lower level showing good
performance and higher percentage showing lower performance. Also if we consider training mode the 10 fold cross
validation is showing significant performance than 66% of split. This results proved that multilayer perception is the
best classifier for the nursery dataset and nave Bayesian is the lowest.
5|Page
Assignment: 1
0.012
0.01
HUNDREDS
HUNDREDS
0.008
0.006
0.004
0.002
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
not_recom
recommend
priority
spec_prior
Naive
Bayesian
0.012
0.01
0.008
0.006
0.004
0.002
0
Multilayer Support
Perceptron Vector
Machine
very_recom
J48
k-nearest
neighbor
not_recom
recommend
priority
spec_prior
Naive
Bayesian
very_recom
The above graphs are showing the precision comparison of five classifier against five classes that we have identified.
Under cross validation training mode the class recommended shows zero precision among all classifier we have used.
But class not_recommeded shows high precision in both training mode. But ver_recommeded class show significant
precision in 66% split. Considering above fact the cross validation again lead in performance.
TP Rate for % split
recommend
priority
spec_prior
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
not_recom
recommend
priority
spec_prior
Multilayer Support
Perceptron Vector
Machine
J48
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
k-nearest
neighbor
Naive
Bayesian
Considering the True positive rate the not recommended class is showing similar results under both testing mode
and all types of classifier used same like us the priority and specific priority class. But the class very recommended is
showing significant different between classifiers and testing mode. However we got good results in multilayer
perception and J48 under cross validation testing mode.
FP Rate for % split
recommend
priority
spec_prior
very_recom
0.12
0.1
0.1
0.08
0.08
0.06
0.06
not_recom
recommend
priority
spec_prior
very_recom
0.04
0.04
0.02
0.02
0
Multilayer Support
Perceptron Vector
Machine
6|Page
J48
k-nearest
neighbor
Naive
Bayesian
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Assignment: 1
False positive rate is high in 66% split rather than cross validation for priority and specific priority class when using
nave Bayesian classifier and this is less in SVM classifier.
Recall for % split
recommend
priority
spec_prior
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
not_recom
recommend
priority
spec_prior
Multilayer Support
Perceptron Vector
Machine
J48
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
k-nearest
neighbor
Naive
Bayesian
Recall measurement shows better performance under cross validation. However very recommended class shows
greater difference between the classifiers. But under 66% percentage split we can see lots of differences in recall.
F-Measure for % split
recommend
priority
spec_prior
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
not_recom
recommend
priority
spec_prior
Multilayer Support
Perceptron Vector
Machine
J48
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
k-nearest
neighbor
Naive
Bayesian
Study of F-measure does not affect significantly in both testing mode and different classifiers except the one class
which is very-recommended.
ROC Area for % split
recommend
priority
spec_prior
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
not_recom
recommend
priority
spec_prior
very_recom
1.2
1
0.8
0.6
0.4
0.2
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
When considering the ROC values the not recommended, specific priority and priority classes are showing high
performance than very recommended and recommended class. In overall view the multilayer perception shows good
performance in classification.
7|Page
Assignment: 1
To analyze the ROC performance the above model was developed.by running the model ROC curves were obtained
for different classifiers of particular class.
Class :- recommended
Class: - priority
When seeing the above ROC curves of classes the recommended class shows poor performance for most of the
classifiers. May due to less amount of instances in that class. (Depend on the data set). Most of the time multilayer
perception gives the good performance. The analysis of ROC time consuming process therefore I did only for cross
validation mode.
8|Page
Assignment: 1
svm
J48
K-nearest neighbor
Nave Bayesian
Conclusion
As a conclusion, we have met our objective which is to evaluate and investigate five selected classification algorithms
based on Weka. The best algorithm based on the nursery data is multilayer perception classifier with an accuracy of
99.7299% and the total time taken to build the model is at 69.56 seconds. When considering the time factor
multilayer perception is more time consuming. According to the time factor k-nearest neighbor and nave Bayesian
classifiers took less time but their accuracy is relatively lower than the multilayer perception. By considering all
aspects of performance parameter under two types of training method the multilayer perception significantly
provide the more accurate results. Also the performance of other classification methods are in decreasing order such
as SVM,J48,k-nearest neighbor and nave Bayesian.
9|Page
Assignment: 1
Classification
Multilayer
Perceptron
Support Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
16.25
97.2699
2.7301
0.7265
0.0319
0.1488
27.9566
63.3737
0.62
94.1498
5.8502
0.0585
0.2419
51.2583
103.0411
0.15
98.0499
1.9501
0.8149
0.0234
0.1336
20.4604
56.914
95.4758
4.5242
0.5306
0.0456
0.2126
39.9548
90.5775
0.01
93.1357
6.8643
0.5385
0.088
0.2271
77.1257
96.7281
10 | P a g e
Assignment: 1
Classification
Multilayer
Perceptron
Support Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
16.21
97.2428
2.7572
0.7522
0.0336
0.1553
29.124
64.7703
0.27
93.8494
6.1506
-0.0005
0.0615
0.248
53.3871
103.4332
0.07
98.807
1.193
0.8943
0.0146
0.1054
12.685
43.9447
96.1824
3.8176
0.6465
0.0384
0.1953
33.3689
81.4648
0.01
92.6034
7.3966
0.5249
0.0888
0.2294
77.0863
95.6866
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
100
98
96
Cross validation 10
66% split
94
92
90
88
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Cross validation 10
Naive
Bayesian
66% split
8
6
4
2
0
Multilayer Support
Perceptron Vector
Machine
J48
Cross validation 10
k-nearest
neighbor
Naive
Bayesian
66% split
Capa Statistics
1
0.8
0.6
0.4
0.2
0
-0.2
Multilayer Support
Perceptron Vector
Machine
J48
Cross validation 10
11 | P a g e
k-nearest
neighbor
Naive
Bayesian
66% split
Assignment: 1
J48
k-nearest
neighbor
Cross validation 10
Naive
Bayesian
66% split
120
100
80
60
40
20
0
Multilayer Support
Perceptron Vector
Machine
RMs
Multilayer Support
Perceptron Vector
Machine
RMs
J48
RAE
k-nearest
neighbor
J48
RAE
k-nearest
neighbor
Naive
Bayesian
RRSE
Naive
Bayesian
RRSE
Above two graphs is showing comparison of different error parameters, considerably the J48 classifier showing
good results since gives lower error rate. Except others, but LIBSVM and Naise bayes show high amount of error
in determining the five classes.
10-fold cross-validation
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Multilayer
Perceptron
12 | P a g e
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Naive
Bayesian
Root relative
squared
error
Assignment: 1
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Multilayer
Perceptron
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Root relative
squared
error
Naive
Bayesian
The above two graphs are showing the compared performance matrices in percentage. The close look of this graph
showing no significant changes between the parameters. The lower level showing good performance and higher
percentage showing lower performance. Also if we consider training mode the 10 fold cross validation showing
significant performance than 66% of split. This results proved that multilayer perception is the best classifier for the
nursery dataset and nave Bayesian is the lowest.
TP Rate
FP Rate
Precision
Recall
FMeasure
ROC Area
Confusion Matrix
0.992
0.333
0.98
0.992
0.986
0.95
0.667
0.008
0.833
0.667
0.741
0.95
a b <-- classified as
3497 44 | a = negative
60 171 | b = sick
0.941
0.97
0.5
0.5
J48(negative)
0.993
0.213
0.987
0.993
0.99
0.878
J48(Sick)
0.787
0.007
0.868
0.787
0.825
0.878
0.984
0.52
0.968
0.984
0.976
0.739
0.48
0.016
0.655
0.48
0.554
0.739
0.94
0.213
0.986
0.94
0.963
0.92
0.787
0.06
0.45
0.787
0.573
0.92
Classification
Multilayer
Perceptron(negative)
Multilayer
Perceptron(Sick)
Support
Vector(negative)
Machine
Support Vector(Sick)
Machine
k-nearest
neighbor(Negative)
k-nearest
neighbor(Sick)
Naive
Bayesian(Negative)
Naive
Bayesian(Sick)
a b <-- classified as
3540 1 | a = negative
231 0 | b = sick
a b <-- classified as
3523 18 | a = negative
27 204 | b = sick
a b <-- classified as
3484 57 | a = negative
87 144 | b = sick
a b <-- classified as
3314 227 | a = negative
52 179 | b = sick
According to the above results we can conclude J48 has the good classification since it has both TP Rate and FP Rate
higher when used percentage split test.
13 | P a g e
Assignment: 1
TP Rate
FP Rate
Precision
Recall
FMeasure
ROC Area
Confusion Matrix
0.988
0.26
0.983
0.988
0.985
0.951
0.74
0.012
0.795
0.74
0.767
0.951
a b <-- classified as
1197 10 | a = negative
25 50 | b = sick
0.939
0.968
0.5
0.5
J48(negative)
0.995
0.117
0.992
0.995
0.994
0.951
J48(Sick)
0.883
0.005
0.919
0.883
0.901
0.951
0.984
0.377
0.976
0.984
0.98
0.806
0.623
0.016
0.716
0.623
0.667
0.806
0.936
0.225
0.985
0.936
0.96
0.925
0.775
0.064
0.441
0.775
0.562
0.925
Classification
Multilayer
Perceptron(negative)
Multilayer
Perceptron(Sick)
Support
Vector(negative)
Machine
Support Vector(Sick)
Machine
k-nearest
neighbor(Negative)
k-nearest
neighbor(Sick)
Naive
Bayesian(Negative)
Naive
Bayesian(Sick)
a b <-- classified as
1207 0 | a = negative
75 0 | b = sick
a b <-- classified as
1198 9 | a = negative
16 59 | b = sick
a b <-- classified as
1188 19 | a = negative
39 36 | b = sick
a b <-- classified as
1135 72 | a = negative
16 59 | b = sick
According to the above results we can conclude J48 has the good classification since it has higher TP Rate higher
when used percentage cross validation 10 fold test.
So finally according to all above classifiers J48 is the good classifier for the sick dataset. Since it has provided better
performance on both cross validation and split percentage.
ROC Curve
14 | P a g e
Assignment: 1
15 | P a g e
Assignment: 1
The above four ROC curve ,we can identify when we use J48 classifier with cross validation(fold 10) testing method for the
above sick datasets the giving better smooth curve it shows the better classifier is J48 out of all the above five classifier.
If we order the classifier according to the all above result it will be like following order(the lowest numer giving higher
performance).
1.
2.
3.
4.
5.
J48
Nave bayes
Multi layer perception
K-nearest neigbour
LibSVM.
Conclusion
Out of all above results in order to analyze the performance of a classifier though J48 classifier gave the better
performance for the sick dataset, it understood different classifier may give better performance for the different
datasets, which means the performance of a classifier depend on number of instances, number of attributes. But
anyhow in order to classify certain data we have to consider higher number of instances and higher number of
attributes. But finally to take the proper decision we have to run the same datasets through using different
classifier and different testing mode such as different values of cross validation and appropriate percentage split
(but 66% is the standard value).
16 | P a g e
Assignment: 1
wine
Instances: 178
Attributes: 14
Class
Alcohol
Malic_acid
Ash
Alcalinity_of_ash
Magnesium
Total_phenols
Flavanoids
Nonflavanoid_phenols
Proanthocyanins
Color_intensity
Hue
OD280/OD315_of_diluted_wines
Proline
Classification
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
Multilayer
Perceptron
0.77
97.191
2.809
0.9574
0.0247
0.1172
5.6355
25.0058
Support Vector
Machine
0.11
98.3146
1.6854
0.9745
0.226
0.279
51.4678
59.5404
J48
0.04
93.8202
6.1798
0.9058
0.0486
0.2019
11.0723
43.0865
94.9438
5.0562
0.9238
0.0413
0.1821
9.3973
38.8682
0.01
96.6292
3.3708
0.9489
0.0217
0.1294
4.9371
27.6176
k-nearest
neighbor
Naive
Bayesian
Classification
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Kappa
statistic
Mean
absolute
error
Root
mean
squared
error
Relative
absolute
error
Root
relative
squared
error
Multilayer
Perceptron
0.74
96.7213
3.2787
0.9506
0.0252
0.128
5.6297
26.5694
Support Vector
Machine
0.06
98.3607
1.6393
0.9753
0.2259
0.2788
50.54
57.8844
J48
86.8852
13.1148
0.8027
0.0874
0.2957
19.5639
61.3956
95.082
4.918
0.926
0.0431
0.1792
9.6393
37.2046
0.01
98.3607
1.6393
0.9753
0.0124
0.0713
2.7794
14.8027
k-nearest
neighbor
Naive
Bayesian
17 | P a g e
Assignment: 1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
J48
k-nearest
neighbor
Naive
Bayesian
Cross validation 10
66% split
100
Cross validation 10
66% split
95
90
85
80
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
66% split
14
12
10
8
6
4
2
0
Multilayer
Perceptron
Support
Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Capa Statistics
Cross validation 10
1.2
1
0.8
0.6
0.2
18 | P a g e
66% split
0.4
0
Multilayer Support
Perceptron Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Assignment: 1
Cross validation 10
66% split
0.2
0.15
0.1
0.05
0
Multilayer
Perceptron
Support
Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
RAE
RRSE
59.5404
51.4678
60
50
43.0865
38.8682
40
27.6176
25.0058
30
20
10
5.6355
0.1172
11.0723
0.279
9.3973
0.2019
0.1821
4.9371
0.1294
Multilayer
Perceptron
Support
Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
RAE
61.3956
57.8844
50.54
60
RRSE
50
37.2046
40
25.5694
30
19.5639
20
10
5.6297
0.128
9.6393
0.2788
0.2957
0.1792
14.8027
2.7794
0.0713
0
Multilayer
Perceptron
Support
Vector
Machine
J48
k-nearest
neighbor
Naive
Bayesian
Above two graphs are showing comparison of different error parameters, considerably the multilayer
perception classifier showing good results that means lower error rate. Except others, but Support vector
Machine and J48 show high amount of error in determining the five classes.
19 | P a g e
Assignment: 1
10-fold Cross-Validation
100%
80%
60%
40%
20%
0%
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Multilayer
Perceptron
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Root relative
squared
error
Naive
Bayesian
80%
60%
40%
20%
0%
Time
taken
seconds
Correctly
Classified
Instances
Incorrectly
Classified
Instances
Multilayer
Perceptron
Kappa
statistic
Support
Vector
Machine
Mean
absolute
error
J48
Root mean
squared
error
k-nearest
neighbor
Relative
absolute
error
Root relative
squared
error
Naive
Bayesian
The above two graphs are showing the compared performance matrices in percentage. The close look of this
graph showing no significant changes between the parameters. The lower level showing good performance and
higher percentage showing lower performance. Also if we consider training mode the 10 fold cross validation
showing significant performance than 66% of split. This results proved that multilayer perception is the best
classifier for the Wine dataset and nave Bayesian is the lowest.
Final Conclusion
Finally, This study focuses on finding the right algorithm for classification of data that works better on diverse
data sets. However, it is observed that the accuracies of the tools vary depending on the data set used. It
should also be noted that classifiers of a particular group also did not perform with similar accuracies. Overall,
the results indicate that the performance of a classifier depends on the data set, then number of instances
especially on the number of attributes used in the data set and one should not rely completely on a particular
algorithm for their study. So, we recommend that users should try their data set on a set of classifiers and choose
the best one.
20 | P a g e
Assignment: 1
References
1. Gopala Krishna, Bharath Kumar and Nagaraju Orsu Performance Analysis and Evaluation of Different Data
Mining Algorithms used for Cancer Classification, (IJARAI) International Journal of Advanced Research in
Artificial Intelligence, Vol. 2, No.5, 2013.
2. Mohd Fauzi bin Othman and Thomas Moh Shan Yau Comparison of Different Classification Techniques
Using WEKA for Breast Cancer IFMBE Proceedings Vol. 15.2007.
3. Rohit Arora and Suman Comparative Analysis of Classification Algorithms onDifferent Datasets using
WEKA, International Journal of Computer Applications (0975 8887),Volume 54 No.13, September 2012
4. Samrat Singh and Vikesh Kumar Performance Analysis of Engineering Students for Recruitment Using
Classification Data Mining Techniques Samrat Singh et al , IJCSET , Vol 3, Issue 2, 31-37 ,February 2013 .
21 | P a g e