Page 172 - ICDEBI2018

I
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456nternational Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470 | IF: 4.101
and accuracy is compared using WEKA data mining and accuracy is compared using WEKA data mining which drawn from training dataset. Classification which drawn from training dataset. Classification
tool. model estimate the likelihood of the record beloodel estimate the likelihood of the record belonging
m
t to each class. The class with highest prevents for Y to o each class. The class with highest prevents for Y to
According to literature review it is found that Naïve eview it is found that Naïve happen when events for X probability becomes the happen when events for X probability becomes the
[2]
Bayes is suitable classification algorithm for multi Bayes is suitable classification algorithm for multi class label for the record.
attribute analysis. It is essential to develop user attribute analysis. It is essential to develop user
friendly application which useful in any education friendly application which useful in any education Definition of Bayes Theorem:Definition of Bayes Theorem: Given two random
sector. Researcher developed application using C# for sector. Researcher developed application using C# for variables X and Y, each of them taking a specific variables X and Y, each of them taking a specific
predicting learning ng behavior behavior of of student student by by value corresponds to a random event. A conditional to a random event. A conditional
implementing Naïve Bayes theorem. probability P(X/Y) represents the probability of robability P(X/Y) represents the probability of
p
events for Y to happen when event for X have already vents for Y to happen when event for X have already
e
[2]
III. Classification Techniques: occurred.
Classification is a supervised learning method where Classification is a supervised learning method where P(X/Y) = P(X/Y).P(Y)
data is divided into different categories or classes. data is divided into different categories or classes. P(X)
The objective of classification to predict target class The objective of classification to predict target class
for given dataset. There are various techniques of given dataset. There are various techniques of P(Y/X) = P(X/Y).P(Y)
classification like decision tree, Naïve Bayes classification like decision tree, Naïve Bayes P(Y)
classifier, nearest neighbor approach, artificial neural classifier, nearest neighbor approach, artificial neural
network network these these are are important important techniques techniques of of V. Training Dataset:
classification. Accuracy of target prediction is classification. Accuracy of target prediction is Following table shows training dataset of MCA I year Following table shows training dataset of MCA I year
depends upon selection of classification technique. In selection of classification technique. In student dataset. Here researcher is interested to predict student dataset. Here researcher is interested to predict
many many real real life life situations situations classification classification is is learning behavior of student from given training learning behavior of student from given training
fundamentally probabilistic, it is uncertain to which fundamentally probabilistic, it is uncertain to which dataset using Naïve Bayes algorithm. Studentdataset using Naïve Bayes algorithm. Student data
[1]
consists of different attributes like Gender, Area, onsists of different attributes like Gender, Area,
class record is belong. c
SSC_Medium, SC_Medium, S H
S
HSC_faculty, SC_faculty,
SSC_Percentage, SC_Percentage,
IV. Naïve Bayes Classifier: M
Math_At_HSC,Graduation_Marks,Admission_Type,ath_At_HSC,Graduation_Marks,Admission_Type,
Bayesian classification is based on Bayes theorem. Bayesian classification is based on Bayes theorem. Entrance_Rank,ParentsIncome,,Attendance,CommuniEntrance_Rank,ParentsIncome,,Attendance,Communi
The posterior probability of the class that a record The posterior probability of the class that a record cation_Skill, Learning_Behavior (Class Label) etc.cation_Skill, Learning_Behavior (Class Label) etc.
belongs to is an approximated using prior probability belongs to is an approximated using prior probability

Table 1: Training Dataset:

@ IJTSRD | Available Online @ www.ijtsrd.comwww.ijtsrd.com | Conference Issue: ICDEBI-2018 | | Oct 2018 Page: 164