You are on page 1of 5

ISBN: 978-15-08772460-24

Date: 8.3.2015
PROCEEDINGS OF INTERNATIONAL CONFERENCE ON RECENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY

EFFICIENCY IMPROVEMENT IN CLASSIFICATION TASKS USING NAIVE BAYES


TREE AND FUZZY LOGIC
Revathi.K*Jawahar.M**
* ComputerScienceand Engineering,
K.S.RInstituteforEngineeringandTechnology,Namakkal,Tamilnadu,
revathisanjucse@gmail.com

**ComputerScienceandEngineering,
K.S.RInstituteforEngineeringandTechnology,Namakkal,Tamilnadu,
mjawahar@gmail.com

ABSTRACT
To improve the classification accuracy rates for Naive Bayestree (NBTREE) and Fuzzy Logic
for multi class problem. In our first proposed NBTREE algorithm, due to presence of noisy
inconsistency instances in the training set its may cause nave bayes classifiers tree suffers from
over fittings its decrease accuracy rates then we have to compute nave Bayes tree algorithm
(NBTREE)to remove the unwanted noisy data from large amount of training dataset. Then our
second proposed fuzzy logic algorithm, we apply nave Bayes tree (NBTREE) to select similarly
more important subset of attributes for the production of naive assumption of class conditional
independence, to improve extract valuable training datasets and We tested the performances of
the two proposed algorithm against those existing systems are Nave Bayestree induction and
Fuzzy logic classification respectively using the classification accuracy validation. Thus result
may cause that identity the most sufficient attributes for the description of instances and also
accuracy rates has been improved.
KEYWORDS: Classification,NaveBayestree(NBTree),Fuzzy Logic, Decision tree
induction,navebayesclassiifers
INTRODUCTION:
Classification is an important task in data
mining .Currently, there are many
classificationproblems where large training
datasets are available,and thereforethere is a
big interest for developing classifiers that
allow handlingthis kind of datasets in a
reasonable time.
It also predicts categorical class
labels and classifies data based on the
training set values further its used to
classifying a new data.This paper examines
the performance of Bayesian networks as
classifiers, comparing their performance to
that of the Nave Bayes(NBtree) classifier

IAETSD 2015: ALL RIGHTS RESERVED

and the Fuzzy Logic classifier, both of


which make strong assumptions about
interactions
between
domain
variables.First,in many cases the accuracy of
nave bayes classifiers is comparable or
higher than the accuracy of other
classification models. Second,most fuzzy
logic do not require a large number of
parameters to be adjusted in their design.
Third,due to their intuitively appealing
topology between nave bayes classifier and
fuzzy logic ,the resulting classification
models become easy to comprehend with
both fuzzy logic and nave bayestree.

www.iaetsd.in

44

ISBN: 978-15-08772460-24
Date: 8.3.2015
PROCEEDINGS OF INTERNATIONAL CONFERENCE ON RECENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY

RELATED WORK
Chandra and Varghese (2009) The G-FDT
tree used the Gini Index as the split measure
to choose the most appropriate splitting
attribute for each node in the decision tree.
Motivated
by
performance
and
comprehensibility
considerations,
we
propose a new node splitting measure in this
paper.We show that the proposed measure is
convex and well behaved. Our results over a
large number of problems indicate that the
measure results in smaller trees in a large
number of the cases without any loss in
classification accuracy.
LEEet.al (2010)
A Nave Bayes
classification enhancing technique, the
ACDDWeighting
Factor
Facility
is
presented and described here. The
enhancement is seen in terms of an
improvement in the classification accuracy
and is achieved through applying unique
weighting factors to each category based on
the number of documents that are annotated
to them. The results from our experiments
show that the ACDD weighting factor
facility has improved the classification
accuracy ofthe ordinary Nave Bayes
classification approach.
LEVENT et.al (2011) The Nave Bayes
method, which is the simplest form of a
Bayesiannetwork, is a popular data mining
method that has been applied to many
domains, including intrusion detection. The
methods simplicity relies on the assumption
that all of the features are independent of
each other. The HNB method, which relaxes
this assumption, has been successfully
applied to web mining.
PROSPOSED WORK:

IAETSD 2015: ALL RIGHTS RESERVED

Propose a new method to improve


the classification accuracy ratesusing Nave
Bayes tree(NBTREE) and fuzzy logic.
Nave Bayes tree(NB TREE)
Fuzzy logic
Nave Bayes tree (NB TREE):The
NBTREE algorithm is
similar to the
classical recursive partitioning schemes, not
including that the leaf nodes created are
naive Bayes categorizers in reverse of nodes
predicting a single class. It attempts to
approximate whether the generalization
accuracy for a nave-Bayes classifier at the
current node.
Fuzzy logic:Fuzzy logic is that it accepts the
uncertainties that are inherited in the
realistic inputs and it deals with these
uncertainties in their affect is negligible and
the resulting in a precise outputs. Fuzzy
logic reduces the design steps and reduced
complexity.
EXPERMENTIAL SETUP:
To test the proposed hybrid methods, we
have used the classification accuracy and
10- fold cross validation.To improve the
classification accuracy rates for Naive
Bayestree (NBTREE) and Fuzzy Logic for
multi class problem
Equation 1 accuracy rates

ANALYZING THE DATA SET

www.iaetsd.in

45

ISBN: 978-15-08772460-24
Date: 8.3.2015
PROCEEDINGS OF INTERNATIONAL CONFERENCE ON RECENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY

A data set is a collection of data,


commonly presented in tabular form. Each
column serves as particular variable. Each
row coincidesto a given member of the data
set in question. It listing the values for each
set of the variables, such as height and
weight of an object of random numbers.
Each value is established as a datum. The
data set may contain data for one or more
members, comparable to the number of
rows.
The attribute Diagnosis was
identified as the predictable attribute with
value 1 for patients with heart disease and
value 0 for patients with no heart disease.
The attribute Patient ID was used as the
key; the rest are input attributes. It is
pretended that problems such as missing
data, inconsistent data, and equivalent data
have all been resolved. Here in our project
we get a data set from .dat file as our file
reader program will get the data from them
for the input of Nave Bayes based mining
process.

Questionnaires have advantages over


some other types of medical symptoms that
they are reasonable, do not lack as much
effort from the questioner as verbal or
telephone surveys, and generally have
standardized answers that make it simple to
compile data. However, such standardized
answers may confront users. Questionnaires
are also clearly limited by the fact that
respondents must be able to read the
questions and respond to them.

Questionnaires is based on the


attributes given in the dataset
Each attribute has separate
questions in terms of values
Questions are given to the
respondents in simple way with
higher understanding
Analyses
is
done
after
enrollment of the answers

FIGURE 2 EXISTING ALGORTHIM

NAIVES BAYES IMPLEMENTATION


IN MINING:
Figure 1 Naive Bayes Classifiers

DESIGNING
ATTRIBUTES

THE

IAETSD 2015: ALL RIGHTS RESERVED

INPUT

Bayes' Theorem finds the probability


of an event occurring given the probability
of another event that has already occurred. If
X represents the dependent event and
Yproduce the prior event, Bayes' theorem
can be stated as follows.

www.iaetsd.in

46

ISBN: 978-15-08772460-24
Date: 8.3.2015
PROCEEDINGS OF INTERNATIONAL CONFERENCE ON RECENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY

Responses are analyzed


Depending upon the responses
graph structure is resulted
Using
Bayes
theorem
techniques
Nave
bayes
classifier is implemented
Accuracy rate is calculated
Comparative analyses on the
obtained result is done further
COMPARISON OF ACCURACY RATES
The obtained result for patient is
compared with the already existing result
and the accuracy is calculated.
RESULT:

Table 2:Fuzzy Logic

NAVE BAYES TREE


The classification accuracies for a Nave
Bayes Tree with 6 fold cross validation.

Figure2 : Graph of Fuzzy Logic

Table 1:Nave Bayes Tree

Figure2: Graph of Nave Bayes Tree

FUZZY LOGIC:
The classification accuracies values
for a Fuzzy Logic with 6 fold cross
validation.

IAETSD 2015: ALL RIGHTS RESERVED

CONCLUSION:
In this, new techniques called nave
Bayes tree (NBTREE) and fuzzy logic is
used for any type of dataset to display the
most accuracy rates for dataset. The
NBTREE and fuzzy logic will makes
preprocessing and pruning tree more
accuracy than the nave Bayes classifiers
and decision tree induction. The comparison
of two dataset will be more accurate than of
existing system.
In this, study is made about domain
knowledge and also the literature survey is
conducted in this the area of classification
techniques and alogrthim.the design of
proposed system is prepared to solve the
problem in the existing system.
REFERNCES:

www.iaetsd.in

47

ISBN: 978-15-08772460-24
Date: 8.3.2015
PROCEEDINGS OF INTERNATIONAL CONFERENCE ON RECENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY

1. Aitkenhead, M. J.(2008) A coevolving decision tree classification


method Expert Systems with
Applications.

Bayesian classification of microarray


data.
Expert
Systems
with
Applications, 37,81888192.

2. Aviad, B. & Roy G. (2011)


Classsification
by
clustering
decision tree-like classifier Expert
Systems with Applications.

10. Franco-Arcega, A.Carrasco-Ochoa,


J.A. Sanchez-Diaz, G. & MartinezTrinidad, J. F. (2011) Decision tree
induction using a fast splitting
attribute selection for large datasets.
Expert Systems with Applications,
38, 1429014300.

3. Balamurugan,S. A .A. &Rajaram, R.


(2009) Effective solution for
unhandled exception in decision tree
induction
algorithms.
Expert
Systems with Applications, 12113
12119.
4. Breiman, L. Friedman, J. Stone, C. J.
&Olshen, R. A. (1984)Classification
and regression trees.

11. Hsu, C.-C., Huang, Y.-P & Chang,


K.-W. (2008) Extended naive Bayes
classifier formixed data. Expert
Systems with Applications, 35,
10801083.

5. Bujlow, T. Riaz,. (2012) A method


for classification of network traffic
established on C5.0 machine
learning algorithm. (pp. 237241).

12. Koc, L. Mazzuchi, T. A &Sarkani, S.


(2012). A network intrusion
detection system based on a hidden
naive Bayes classifier. Expert
Systems with Applications, 42,
1349113500.

6. Chandra, B. & Gupta, M. (2011)


Robust approach for estimating
probabilities in naive Bayesian
classifier for gene expression data.
Expert Systems with Applications,
(pp.12931298).

13. Lee, L. H. & Isa, D. (2010).


Automatically computed document
dependent weighting factor facility
for nave Bayes classification.
Expert Systems with Applications,
37, 84718478.

7. Chandra, B., & Paul Varghese, P.


(2009) Moving towards efficient
decision
tree
construction.
Information Sciences, 179, 105910
8. Chandra, B. & Varghese, P. P.
(2009) Fuzzifyinggini index based
decision trees. Expert Systems with
Applications, 36, 85498559.

9. Fan, L. Poh, K.-L. & Zhou, P. (2010)


Partition-conditional
ICA
for

IAETSD 2015: ALL RIGHTS RESERVED

www.iaetsd.in

48

You might also like