You are on page 1of 6

IPASJ International Journal of Computer Science (IIJCS)

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm


A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

An Innovative Technique to Improve


Students Performance with Neural Network
and Data Mining Technique
Nikhil Sunil Wadhwani1 Amol Joglekar2
1
Dept. of Computer Science
Mithibai College, Mumbai, India
2
Dept. of Computer Science
Mithibai College, Mumbai, India

ABSTRACT
ANN stands for Artificial Neural Network and is a computational model. It takes into account all the biological functions.
They are used in system identification, vehicle control, process control, natural resources management, game-playing and
decision making, pattern recognition, sequence recognition, medical diagnosis, financial applications, data mining for student
performances, visualization, e-mail spam filtering, etc. With the help of various factors like health, habits, past scores, etc. of a
student, we can estimate and help students improve future performances.
Keywords: Artificial Neural Network, Student Performances, Education, Big Data, Predictive Model

1. INTRODUCTION
Data mining techniques were first applied in various domains like Banking, Communication, Engineering, etc. Now
we can use Data mining in nearly everything including tasks related to Education. With Data mining and Artificial
Neural Networks, we not only can predict a students current growth but can customize his efforts in a way that can
help him build his future. Neural Networks take into account all scenarios like past performance, participation in co-
curricular activities, peer groups, social activities, health status, financial background, family conditions, etc. Knowing
these factors can help us improve and boost a students performance.
The main objective behind the proposed work is to check and analyze various factors that can affect a students
performance, transforming them into a model that can help students better their skills.

1.1 Neural Network


The idea of Artificial Neural Network (ANN) is based on the belief that, by making the right connections, working of
the human brain can be imitated using silicon and wires, as living neurons and dendrites. Stimuli from external
environment or inputs from sensory organs are accepted.
These inputs create electric impulses, which quickly travel through the neural network. A neuron can then send the
message to other neurons to handle the issue or not forward it.
ANN consists of multiple nodes. These nodes act like an entrance to the human brain. Each of these nodes gets
connected via links. The nodes can take input data and perform the needed operations on the data. The result of these
operations is passed to other neurons which give us an output called its activation or node value. Each link is associated
with weight. ANNs are capable of learning, which takes place by altering weight values.

1.2 Big Data


Big data is a term used in databases and warehouses where a huge lump of data needs to be stored in a database or a
warehouse.
The data lying in the servers of company was just data until yesterday sorted and filed. The data in your company is
none other than Big Data. Big data is all data the company ever stores.
It includes data stored in clouds and even the URLs that you bookmarked. All the data present in servers are
collectively called big data.

Volume 5, Issue 10, October 2017 Page 44


IPASJ International Journal of Computer Science (IIJCS)
Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm
A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

All this data can be used to get different results using different types of analysis. Based on the results and predictions,
big data is used accordingly. Big Data is essentially the data that you analyze for results that you can use for predictions
and for other uses.
When using the term Big Data, there are many similar data sets exists. But it is a boon to have replication in the big
data market.

1.3 Predictive analysis


Predictive analysis uses various statistical methods that range from predictive modeling to machine learning and data
mining.
It analyzes current and historical facts that help make predictions about future or any event. In companies, we get many
patterns over a period of time. These patterns help to predict future patterns.
By predicting, the company can assess the risks or potential risks associated by a set of conditions and make decisions
that are conducive for the companys growth.
In order to get next to accurate results statistics along with predictive analysis is used.

Figure 1: Working of Artificial Neural Network.

2. Literature Review
Bhardwaj and Pal [2] used the Bayesian classification methods based on various attributes. A Study was made on 300
students from 5 different degree colleges conducting BCA (Bachelor of Computer Application).
Factors like living location, teaching medium, family status, etc. were highly correlated with a students academic
performance.
Vuda Sreenivasa Rao [1] studied the concepts of soft-computing and data warehousing. Using data mining a model
for improving the performance of students was created.
Mukta and Usha [3] used and studied the technique of Artificial Neural networks along with old statistical methods.
The results were compared and performance was evaluated for students of a business school. This helped predict the
academic performance of those graduates.
Kanakana and Olanrewaju [4] collected data from Tshwane University of Technology. They used Artificial Neural
Network and linear regression models to predict student performance. An average point score was taken as the input
variable.
The results indicated a better agreement between ANN model prediction and observed values compared to those in the
linear regression.
Pandey and Pal [5] used Bayes Classification based on categories, languages and background qualification to study
student performances.
A study for the same was conducted by selecting 600 students from various colleges in India that proved whether new
students will perform or not.

Dr. Yashpal Singh and Alok Singh Chauhan [8] talked about the allocation of students by considering a set of clusters
and boundaries in each cluster. By means of Genetic Algorithm, they were able to successfully reduce optimization
problems in allocation.

Volume 5, Issue 10, October 2017 Page 45


IPASJ International Journal of Computer Science (IIJCS)
Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm
A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

3. METHODOLOGY
The proposed work would be conducted using some of the online tools that take down big data in an appropriate
format like excel sheet. After collecting the data, the data is analyzed in a way to create a function that can check
inputs and generate outputs. All data entered is taken as inputs and the resultant data is given as output. For this
algorithm, there is a need of input and output variables.

3.1 INPUT VARIABLES


Table 1: Input Variables
Sr. No. Input Variables Domain
Male
1 Students Gender Female
Other
More than 70%
61-70%
2 Students SSC score
51-60%
Less than 50%
More than 70%
61-70%
3 Students HSC score
51-60%
Less than 50%
Less than 1 Lakh
1-3 Lakhs
4 Annual Family Income 3-6 Lakhs
6-10 Lakhs
10 Lakhs and up
1 Year
1-2 Years
5 Gap in Studies 2-3 Years
More than 3 Years
None
Less than 30 mins
Travelling time to 30 mins - 1 hour
6
College 1 hour - 2 hours
More than 2 hours
Yes
7 Certifications Done
No
Dancing
8 Co-Curricular Activities Singing
Others
1 daily
2 daily
Number of cigarette
9 3 daily
packs consumed
4 or more
Do not smoke
Once
Number of times alcohol Twice
10 is consumed by a student Thrice
in a day More than 4 times
Do not consume
11 Attentiveness in lecture Scale from 1-10
Less than an hour
Amount of time spent on 1 hour- 2 hours
12
social media 2 hours- 3 hours
More than 3 hours
Less than 1 hour
Amount of time spent 1-2 hours
13
gaming or on P.C.s 2-3 hours
More than 3 hours
A day before
Time taken before A week before
14
preparing for any exam A month before
Two months or more
Yes
15 Any disabilities
No

Volume 5, Issue 10, October 2017 Page 46


IPASJ International Journal of Computer Science (IIJCS)
Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm
A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

3.2 OUTPUT VARIABLES


The output variables focus on of how a student is coping and what can be done to improve his/her performance in
future exams.
Table 2: Output Variables Calculation
Sr. No. Result Output
Variable
1. Good 60% or more
2. Poor Less than 60%

4. Proposed Algorithm
Artificial neural network is based on the topology of feeding forward with no back loops. The same is implemented in
the research work by using the Feed-Forward topology with Artificial Neural Networks resulting in the Feed-Forward
artificial neural network. As it is an input to output flow there should be no back loops. There are no limitations on the
number of layers, type of transfer function used in individual artificial neuron or number of connections between
individual artificial neurons [5].

A simple algorithm is implemented in the research that takes in all inputs and generates outputs. All the data goes in
one flow and has no back-loops as required in an Artificial Neural Network. Also, the data is stored in a large database.

The algorithmic steps are as follows:

Step 1. Admin opens the GUI and enters all data that was recorded using Google forms and Google spreadsheet.

Step 2. The upstream sources gives us the input variables i and weights w which are analyzed.

Step 3. Weights and inputs together are stored in a large database.

Step 4. The database stores the big data and checks for any error or malfunction.

Step 5. Positive and Negative points (inputs) are recorded along with their weights.

Step 6. Using the weights we calculate the output layer (Oi).

Step 7. Oi is calculated using steps 4, 5 and 6 with the formula

where Ii and Wi, are the inputs and weights of each input respectively.

Step 8. Admin analyses the output layer function and thus generates results.

Step 9. Results help students improve their performances.

Given the following algorithm we can propose a flowchart that can define the flow of the program.
The flowchart below makes it easier to understand how we can implement ANN ( Artificial Neural Network) with big
data.

Volume 5, Issue 10, October 2017 Page 47


IPASJ International Journal of Computer Science (IIJCS)
Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm
A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

Figure 2: Proposed Working Model

Volume 5, Issue 10, October 2017 Page 48


IPASJ International Journal of Computer Science (IIJCS)
Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm
A Publisher for Research Motivation ........ Email:editoriijcs@ipasj.org
Volume 5, Issue 10, October 2017 ISSN 2321-5992

5. Conclusion and Future Scope


An Artificial Neural Network for predicting student performance model was conceived using feed forward algorithm
for training. The factors for the model were obtained from students who filled the online form. This study showed the
potential of the artificial neural network for predicting student performance.
With the following research we can see that it has a scope to improve. The improvement can be bought by introducing
Artificial Intelligence to the acquired big data in such a way that the student gets instant response and ideas of how to
boost his performance.

References
[1] Sreenivasa Rao& Yohannes (2012) Improving Academic Performance of Students of Defense University Based on
Data Warehousing and Data Mining, Global Journal of Computer Science and Technology. 12(2), 201-209.
[2] B.K Bharadwaj and S. Pal. Data Mining: A prediction for performance improvement using classification,
International Journal of Computer Science and Information Security (IJCSIS), Vol. 9, No. 4, pp. 136-140, 2011.
[3] P.Mukta and A.Usha, A study of academic performance of business school graduates using neural network and
statistical techniques, Expert Systems with Applications, Elsevier Ltd., vol. 36, no. 4, (2009).
[4] G. Kanakana1 and A. Olanrewaju, Predicting student performance in Engineering Education using an artificial
neural network at Tshwane University of Technology, ISEM 2011 Proceedings, (2011) September 21-23,
Stellenbosch, South Africa.
[5] U. K. Pandey, and S. Pal, Data Mining: A prediction of performer or underperformer using classification,
(IJCSIT)
International Journal of Computer Science and Information Technology, Vol. 2(2), pp.686-690, ISSN: 0975-9646,
2011.
[6] https://en.wikipedia.org/wiki/Artificial_neural_network
[7] S.N.Sivanandam and S.N.Deepa, Principles of Soft Computing, Wiley India Edition, 2007
[8] Dr. Yashpal Singh, Alok Singh Chauhan Neural Networks in Data Mining, Journal of Theoretical and Applied
Information Technology, 2005-2009
[9] V.O. Oladokun, A.T. Adebanjo, O.E. Charles- Owaba, Predicting Students Academic Performance using
Artificial Neural Network: A Case Study of an Engineering Course, The Pacific Journal of Science and
Technology, Volume 9. Number 1. May-June 2008

Volume 5, Issue 10, October 2017 Page 49

You might also like