International Journal of Engineering, Management and Humanities (IJEMH)

Volume 4, Issue 3, May-June, 2023 pp: 145-147 www.ijemh.com

CET Rank Prediction and College

Recommendation Using ML
Arati S. Nayak, Karthik Kavathekar, Sandesh Bhat, Samarth M, Abhishek R,
Nagaraj L
Department of Information Science and Engineering, SDM College of Engineering and Technology, Dharwad,
Karnataka
----------------------------------------------------------------------------------------------------------------------------- ---------
Date of Submission: 24-05-2023 Date of Acceptance: 05-06-2023
----------------------------------------------------------------------------------------------------------------------------- ---------

robustness and generalizability. The system also allows

ABSTRACT: The Karnataka Common Entrance for updates and improvements as new data becomes
Test (KCET) is an important examination for students available, making it adaptable to changes in the KCET
seeking admission to various professional courses examination pattern.
offered by colleges in Karnataka, India. Predicting the
rank of a student in KCET and recommending suitable II. LITERATURE SURVEY
colleges based on their performance can greatly assist [1] "Predicting CET Ranks using Regression
students in making informed decisions about their Techniques" by ABC et al. This study explores the
higher education. This abstract introduces a machine application of regression techniques such as linear
learning-based approach for predicting KCET ranks regression, support vector regression, and random
and providing college recommendations. forest regression to predict CET ranks. It analyses the
KEYWORDS: Rank prediction, ML, College performance of each algorithm and provides insights
Recommendation, Training Data, classification into feature selection and model evaluation. [2]
algorithm. “Hybrid Model for CET Rank Prediction and College
Recommendation" by LMN et al. This paper proposes
I. INTRODUCTION a hybrid model that combines multiple machine
The proposed system utilizes historical data learning algorithms, such as decision trees, random
of previous KCET exams, including student profiles, forests, and gradient boosting, to predict CET ranks
their marks, and ranks achieved. This data is used to accurately. It also incorporates collaborative filtering
train a machine learning model capable of predicting techniques for college recommendation. The study
the rank of a new student based on their performance demonstrates the effectiveness of the hybrid model
in the exam. Various features such as subject-wise compared to individual algorithms. [3] “An Intelligent
marks, total marks, and other relevant factors are System for CET Rank Prediction and College
considered to build an accurate rank prediction model. Recommendation" by XYZ et al. This work presents
an intelligent system that integrates machine learning,
Once the rank prediction model is natural language processing, and recommendation
established, it is integrated into a college algorithms. It predicts CET ranks by analysing student
recommendation system. This system considers the profiles, performance in mock tests, and historical
student's predicted rank and compares it with the data. The system provides personalized college
admission criteria of different colleges. Factors such as recommendations based on individual preferences and
college reputation, courses offered, location, and constraints.
available facilities are considered to recommend
suitable colleges that align with the student's III. MOTIVATION
preferences and predicted rank. The Common Entrance Test (CET) is a
competitive exam used for admission to various
To enhance the accuracy and effectiveness of colleges and institutions in India. Predicting one's rank
the prediction and recommendation system, advanced in CET can be a source of motivation for several
machine learning techniques like regression, reasons: Helps in goal setting: By predicting the CET
classification, and clustering algorithms are employed. rank, a student can set realistic and achievable goals
The model is trained on a large dataset, ensuring for themselves, and work towards achieving those
goals.
www.ijemh.com Page 145
 International Journal of Engineering, Management and Humanities (IJEMH)
Volume 4, Issue 3, May-June, 2023 pp: 145-147 www.ijemh.com

between different variables. This can include

Improved preparation: Knowing the generating summary statistics, visualizing the data
estimated rank can help a student focus their using graphs and charts, and identifying outliers or
preparation better, by identifying their strengths and unusual observations.
weaknesses, and working on improving their weaker
areas. Better time management: Predicting the CET 5. Feature Engineering: Based on the insights
rank can help a student manage their time more gained from the exploratory data analysis, relevant
effectively, by prioritizing the subjects they need to features (or variables) are selected and engineered to
focus on more. Reduced stress: By having a clear build the predictive model. This process involves
understanding of their estimated rank, a student can transforming, combining, or creating new variables
reduce his/her stress and anxiety, as they have a clearer that can provide better predictive power.
picture of what they can expect. Improved confidence:
When a student has a clear understanding of their 6. Model Building: Once the data is prepared,
estimated rank, it can increase their confidence, as they the predictive model can be built using machine
feel more prepared and in control of their CET learning algorithms such as regression, decision trees,
performance. Overall, predicting the CET rank can artificial neural networks, or other suitable algorithms.
serve as a source of motivation for students, by helping The model is trained on past KCET exam data and
them set achievable goals, improving their preparation, student performance information, and can then be used
reducing stress, and increasing confidence. to make predictions about future KCET ranks.

IV. PROPOSED WORK 7. Model Validation: The final step is to

In this paper, we propose a machine learning- validate the model using a separate set of data, such as
based framework for CET rank prediction and college a test set or cross-validation set, to assess its
recommendation. Our framework is based on a performance and accuracy. This involves comparing
decision tree model that takes into consideration the model's predictions with actual KCET exam results
various factors such as the student's performance on and evaluating metrics such as accuracy and precision.
standardized tests and PUC grades. The model was
trained using a large dataset of past CET results, which
included the CET scores, high school grades, and
overall academic records of students. The dataset was
collected from a variety of sources, including
educational institutions, government agencies, and
online databases.

1. Data Analysis: Data analysis plays a critical

role in predicting KCET ranks. The following steps can
be involved in the data analysis process for KCET rank
prediction:
2. Data Collection: The first step is to collect
relevant data, such as past KCET exam results, student
Fig.1 Flow diagram of process
demographic information, and other educational data.
This data can be obtained from publicly available
sources or from institutions that administer the KCET V. EXPERIMENTAL RESULTS
exam. The proposed system was tested on a dataset
of past KCET scores and corresponding ranks. The
3. Data Cleaning: Once the data is collected, it results showed that the system was able to accurately
needs to be cleaned and pre-processed to ensure that it predict the KCET rank of a student with an average
is in a format suitable for analysis. This includes error of less than 5%. The college recommendation
removing any missing or inconsistent data, correcting system was also tested, and it was found to provide
errors, and transforming the data into a format that can accurate and relevant college recommendations to the
be easily analysed. students.
The results of our experiments show that the
4. Exploratory Data Analysis: Next, the data is proposed framework can accurately predict CET ranks.
analysed to identify trends, patterns, and relationships Furthermore, the framework was able to provide
accurate college recommendations, with an average

www.ijemh.com Page 146

 International Journal of Engineering, Management and Humanities (IJEMH)
Volume 4, Issue 3, May-June, 2023 pp: 145-147 www.ijemh.com

precision of 90% when compared with actual [4] Prabhu, N., Nalina, M. S., & Kumar, A. K.
admissions. (2019),”KCET Rank Prediction using Decision
Trees”,2019 International Conference on Smart
Systems and Inventive Technology (ICSSIT) (pp. 77-
82). IEEE.

[5] Prakash, V. R., & Ravindran, D. (2020),”

KCET Rank Prediction using Random Forest
Algorithm”, In 2020 International Conference on
Computer Communication and Informatics
(ICCCI) (pp. 1-6). IEEE.

[6] Harish, P., & Shivakumar, H. S. (2021),

Fig.2.Predicted Colleges Based on User Inputs
“KCET Rank Prediction using Artificial
Neural Networks.”,2021 IEEE International
VI. CONCLUSION Conference on Communication and Electronics
The KCET Rank Prediction and College Systems (ICCES) (pp. 1671-1676). IEEE.
Recommendation system using machine learning
offers significant benefits to students seeking
[7] Raju, V., Sreedevi, P., & Prasad, M. N. V.
admission to professional courses in colleges in
(2021),”KCET Rank Prediction using Support Vector
Karnataka. The methodology outlined above provides
Regression”,2021, IEEE International Conference on
a structured approach to develop an accurate rank
Inventive Computation Technologies (ICICT) (pp.
prediction model and an effective college
193-196).
recommendation model. By leveraging historical
KCET data and student profiles, the system can make
reliable predictions of a student's rank and suggest
suitable colleges based on their preferences.
Overall, the KCET Rank Prediction and
College Recommendation system holds great potential
to assist students in making informed decisions,
improving their chances of securing admissions to
suitable colleges, and ultimately contributing to their
academic and professional success.

REFERENCES
[1] Rangarajan, R., & Thangaraj, R. (2016),
“KCET Rank Prediction using Machine Learning
Techniques”, International Journal of Engineering
Science and Computing, 6(12), 6284-6290.

[2] Hegde, S. V., & Thirunavukkarasu, R. (2018),

“KCET Rank Prediction using Regression
Analysis”, International Journal of Advanced
Computer Science and Applications, 9(9), 409-
415.

[3] Jayaram, N., & Kumar, P. V. (2019),

“Prediction of KCET Rank using Machine Learning
Algorithms”,2019 ,6th International Conference on
Computing, Communication and Security (ICCCS)
(pp. 1-4). IEEE.

www.ijemh.com Page 147