8 V May 2020

http://doi.org/10.22214/ijraset.2020.5466
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429
Volume 8 Issue V May 2020- Available at www.ijraset.com

Students Placement Prediction using Machine

Learning
Mr. C K Srinivas1, Nikhil S Yadav2, Pushkar A S3, R Somashekar4, Sundeep K R5
1, 2, 3, 4, 5
Computer Science Department, Visvesvaraya Technological University

Abstract: Placement of students is one of the most important objective of an educational institution. Reputation and yearly
admissions of an institution invariably depend on the placements it provides it students with. Institutions make great efforts to
achieve placements for their students .This will always be helpful to the institution. The objective is to predict the students getting
placed for the current year by analyzing the data collected from previous years students. This model is proposed with an
algorithm to predict the same. The data has been collected by the institution for which prediction is going to be done and by
applying suitable data pre-processing techniques. This model is prepared by using Logistic Regression algorithm. This algorithm
independently predict the results and we then compare the efficiency of the algorithm, which is based on the dataset. This model
will helps the placement cell to focus on the potential students and help them to improve their technical and social skills.
Keywords: Placement Prediction, Logistic Regression, Dataset, Machine Learning, Accuracy, Probability.

I. INTRODUCTION
Every college put their efforts for Placements which is very important for each and every student. The achievement of the college is
known by the number of students getting recruited in their Institutions. Every student takes admission to the colleges by going
through the history of placements in the college. The main approach of the project is collect the historical data of pervious years
students from institution and find the probability of current year students getting placed. This model helps the college to improve
their campus placements. In students placements Prediction model ,it predicts the probability of a student getting placed or not.
This is done by using Logistic Regression as the algorithm. This model is design to predict whether
a particular student getting placed or not in campus recruitment .To check the prediction the data collected from student are
percentage, arrears , cgpa and their domain knowledge.

II. LITERATURE SURVEY

Sr.No Paper name Authors Technique Advantages & Disadvantages
1 Placement Prediction JoshitaGoyal, Shilpa naïve bayes and Advantages:
Decision Support System Sharma improved naïve It used for exploring
Using Data Mining. bayes algorithm the unique types of data in college.
Disadvantages:
It makes assumptions on the data provided, for any
two independent features which are given for the
output class
2 Adaptive Model For Subitha Sivakumar, Improved Decision Advantages:
Campus Placement Rajalakshmi Selvaraj Tree For segregating the eligible students for campus
Prediction Using Improved algorithm selection based on academic performance measures
Decision Tree Disadvantages:
A small change in the data can cause a large
change in the structure decision tree causing
instability.
3 A Data Mining Model To Ravi Tiwari And Awadesh 1)Fuzzy Logic Advantages:
Improve Placement”. Kumar Sharma K Nearest It improve the student’s performance, a work
2)Neighbor has been analyzed and predicted using the
Classification algorithms Fuzzy logic and the KNN algorithm
algorithms Disadvantages:
Fuzzy logic and KNN algorithm is mostly
work on integer value. hard to classify string
type data.

©IJRASET: All Rights are Reserved 2771

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429
Volume 8 Issue V May 2020- Available at www.ijraset.com

4 Predicting Student’s D. Satish Kumar, Zailan Logistic Regression Advantages:

Campus Placement Bin Siri, D.S. Rao, S. Algorithm Predicts based on the individual performances
Probability using Binary Anusha of the students at the time of interview.
Logistic Regression Disadvantages:
when
the academic performance is a bit low student are
not
selected which is unfair.

III. PROBLEM STATEMENT

The placement prediction model considers only academic performances of the students so that the prediction of the student getting
placed or not can be done. We cannot consider the placement of students just by their academic performances because some
students may be good at aptitude, technical and communication skills due to their low score in their academic that may tend to be
their drawback. For predicting the placement of a Student needs parameters like cgpa, logical and technical skills Academic
performances may be important but the model is design to predict the placements based on the parameters of the students.

IV. EXISTING SYSTEMS

The present system generally consider academic performances as a single parameter to judge whether a student can be placed or not
during the campus placements. generally the parameter used to judge the capability of the student, performance in the academics
during the first three years of engineering .Academic performance is not only important for getting selected in an interview but also
depends on the awareness of student during the aptitude tests and interviews. For calculating the probability of a student getting
placed by some data mining algorithms, sometimes gives a probability of more than 100% which is not feasible and denotes a
wrong interpretation to the student .Negative probability is shown from certain algorithms which gives an wrong interpretation to
the student. Academic performance is not only the parameter for judging the student. But other parameters like aptitude and
technical knowledge should be also considered in order to determine the outcome for the student’s future.

V. METHODOLOGY
1) Algorithm: Logistic Regression
Logistic Regression is a one the Machine Learning algorithm which is used for the problems classification, it is a predictive
analysis algorithm and based on the concept of probability. For predicting a value based on history of data, it is necessary to train
the prediction model. In our case, we choose logistic model as a predictor system so we shall be training logistic model using dataset
collected by the institution.Through Logistic Regression algorithm we are able to predict the probability of the students getting
placed or not and display the result in terms of percentage .The data collected by the college is divided into training and testing
dataset which is given to the model for training. This algorithm trains the model to predict the probability of the student getting
placed by the training dataset provided.

Fig 1 Graph for Training set

©IJRASET: All Rights are Reserved 2772

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429
Volume 8 Issue V May 2020- Available at www.ijraset.com

The above graph shows the plot for the two of the parameters we have considered for the prediction .It gives the clear view about
the number of should students getting placed or not by indicating 0 and 1. The decision boundary in the above graph specify the
criteria for the students placement such as specifying the condition for the training model when the quants and cgpa of the students
match then it places the data in that respective position as shown in the above graph .The graph is plotted by taking quants score and
the average cgpa of the students when the model is given with this training set the graph plot as show above.

VI. PROPOSED SYSTEM

We have designed a models for checking prediction of student getting placed or not in a campus drive this model will help both
student and institutions for preparing well in advance for campus recruitment. The main objective of this model is to know the
capability of the student and where he stands by predicting the probability of getting placed. This also help the students and
institutions to improve performance of the potential students. This model will consider the academic history of the student such as
percentage as well as their skills like coding ability, verbal skills, technical skills and aptitude which are tested by companies.
We used logistic regression algorithm on the students data gathered from the institution of previous year. This model take scores of
student in secondary education, average scores of all the semesters in the technical education till date and also some parameters
which adds weightage to kick start the career.

Fig 2 Flowchart of Proposed system

©IJRASET: All Rights are Reserved 2773

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.429
Volume 8 Issue V May 2020- Available at www.ijraset.com

Firstly the project object is designed and the data is collected by the college, after collecting the data we prepare the data into
training and testing datasets. The data so formed is separated into dependent variables and independent variables then the outliers
are removed.
1) The training data is provided to the model design for the training results
2) The model is fine tuned and the training data is further classified according to the parameters specified to the model
3) After the training data is ready the testing dataset which was stored earlier is further processed to the model for the prediction.
4) If the testing values comes out to be true the values are finally set as the prediction
5) If the testing values comes out to be false than the values are again sent back for the fine tuning of the model.

VII. RESULTS
The proposed system results in the probability of the students getting placed in the campus drives. The ‘STUDENTS
PLACEMENT PREDICTION USING MACHINE LEARNING’ provides the help for both students and the institution .The
institution can focus on the potential students by knowing the prediction of this model. The technique we used is the logistic
regression which gives the accuracy of students getting placed . The project mainly gives the information about the students
probability of getting placed in the campus drives which benefits both the students and the institution.

VIII. CONCLUSION
We propose a model to predict the percentage of student getting placed. We have design this model by analyzing the data collected
from the institution of previous years students. We used Logistic Regression Algorithm. Logistic Regression algorithm is applied on
previous year data & current students data to predict the student getting placed in various companies. here we check the eligibility of
candidate on basis of percentage & other technical knowledge .This will help students and institution to improve performance of the
potential students.

IX. ACKNOWLEDGMENT
Salutations to our beloved and highly esteemed institute, “Ballari Institute ofTechnology &Management” for having well qualified
staff and labs furnished with necessary equipment.
I express my sincere thanks to my guide Mr. C K Srinivas for giving me constant encouragement, support and valuable guidance
throughout the course of the project, without whose stable guidance this project would not have been achieved and I would thank
our project coordinators Mr. Jagadish R M and Mr. AzharBiag.
I express wholehearted gratitude to Dr. R.N. KULKARNI who is our respectable HOD of Computer Science Dept. I wish to
acknowledge his help who made our task easy by providing with his valuable help and encouragement.
And also my due thanks to Dr. V.C.PATIL, the principal, as I consider myself very lucky to have such excellent computing facilities
and their inspiration throughout our professional course.
I also thank the non-teaching staff of CS department who guided at the time of difficulties

REFERENCES
[1] Joshita Goyal, Shilpa Sharma,"Placement Prediction Decision Support System using Data Mining" Volume 4 - Issue 2 ( 630 - 635 ) - March - April 2018,
International Journal of Engineering and Techniques (IJET) , ISSN: 2395 - 1303
[2] Subitha Sivakumar and Rajalakshmi Selvaraj, 2017.Adaptive Model for Campus Placement Prediction using Improved Decision Tree. Journal of Engineering
and Applied Sciences, 12: 6069-6075. Asian Journal of Information Technology, 18: 250-260.
[3] Ravi Tiwari and Awadesh Kumar Sharma paper titled” A DATA MINING MODEL TO IMPROVE PLACEMENT”.International Journal of Computer
Applications © 2015 by IJCA Journal Volume 120 - Number 12Year of Publication: 2015
[4] D. Satish Kumar, Zailan Bin Siri, D.S. Rao, S. Anusha“PREDICTING STUDENT’S CAMPUS PLACEMENT PROBABILITY USING BINARY LOGISTIC
REGRESSION”.
[5] Manoj K Shukla et al. “Study on Students Placement Prediction Model Using Logistic Regression”, International journal of research and technology, Vol.05(1),
pp.1-4,2017.
[6] Ajay Shiv Sharma, Swaraj Prince, ShubhamKapoor, KeshavKumar,”PPS - Placement Prediction System using Logistic Regression” in 2014 IEEE International
Conference on MOOC, Innovation and Technology in Education.

©IJRASET: All Rights are Reserved 2774