Basic Details of the Team and
Problem Statement
THEME: Development Of Education
PROBLEM STATEMENT TITTLE: STUDENT
DROPOUT PREDICTION BY USING MACHINE
LEARNING
TEAM NAME: MISSION_OF_404
TEAM LEADER NAME: MOHAN.K
TEAM QUARTINATERS: R . ANUNANDHANA
S.ISHRA
INSITUTE NAME: Christian college of
engineering and technology
Introduction Of Our project
The ML project aims to predict students' dropout and academic success using the semi
structured CSV dataset. The dataset provides a comprehensive collection of attributes
related to students' demographic information, academic performance, and various
1
other factors that may influence their educational outcomes.
By using various machine learning algorithms and techniques, the project seeks to
develop models that can accurately predict whether a student is likely to drop out or
achieve academic success based on the provided data. The predictive models will
utilize a range of features, including student demographics (such as age, gender, and
ethnicity), socioeconomic factors, educational background, study habits,and
performance in previous assessments.
IMPORTANCE OF THE DROPOUT PREDICTION
For What? Student dropout prediction aims to identify students at risk of dropping out, for proactive
intervention and improved retention.
Why? Student dropout prediction helps identify students at risk of leaving early, allowing for
interventions to improve their chances of success.
How? Student dropout prediction uses data to spot students at risk of leaving early
Tamil Nadu dropout % Tamil Nadu (9%)
INDIA DROPOUT % India (12.6%)
GLOBAL % The global student dropout rate is estimated to be around 42%
USING ALOGIRTHMS
KNN Naive Bayes
classification and regression 6 1 Classification, and calculates the
probability of a given input
problems.
preceptron Logistic regression
computations to
detect features in the input data
5 2
two values such as true/false,
yes/no,
support vector Random forest
classifier 4 3 classification and regression
classification, regression and
problems.
outliers detection
Primary objectives:
1 Data Preprocessing:
Perform thorough data cleaning, handling missing values, and transforming features to ensure the dataset's quality
and compatibility with machine learning algorithms.
2 Feature Selection:
Analyze the dataset to identify the most relevant features that strongly correlate with dropout and academic
success. Select a subset of features to train the models and improve computational efficiency.
3 Model Training:
Employ various machine learning algorithms, such as logistic regression, decision trees, random
forests, or gradient boosting, to train predictive models on the processed dataset.
4 Model Evaluation:
Evaluate the trained models using appropriate evaluation metrics, such as accuracy, precision, recall, or
F1-score. Assess the models' performance and compare their results to determine the most effective
approach
Primary objectives:
5 Hyper parameter Tuning:
Optimize the selected models by tuning hyper parameters through techniques like grid search or
random search to enhance their predictive capabilities further.
6 Prediction and Interpretation:
Utilize the best-performing model to make predictions on new, unseen data. Interpret the
model's predictions and analyze the most influential features that contribute to dropout or
academic success.
7 Report and Visualization:
Summarize the project findings, including model performance, feature importance, and insights gained from the
analysis. Visualize the results through charts, graphs, and other appropriate visual representations. Ultimately, the
project aims to provide valuable insights into the factors affecting students' dropout rates and academic success,
allowing educators and policymakers to intervene and implement targeted strategies to support struggling students
effectively.
ACCURACY IN %
KNN
0.86%
6 1 Naive Bayes
0.85%
Perceptron Logistic regression
0.89% 5 2 0.91%
support vector classifier Random forest
0.92%
4 3 0.92%
BENEFITS
Early Intervention
1 By identifying students at risk of
dropping out, schools can
Data-driven Decision intervene early
Making
data-driven helps schools make 5 2 Improved Resource
informed decisions about Allocation:
educational policies and Schools can prioritize resources
interventions. towards at-risk students
Proactive Outreach
Schools can proactively reach
out to students at risk
4 3 Personalized Learning
educators to tailor learning approaches to
individual needs, fostering a more
engaging learning experience.
DOMAINS
Socioeconomic
Domain
Predicting Dropout Risk Based on
Socioeconomic Background
Social Domain
Impact of Peer Networks and Behavioral Domain
Support Systems on Dropout
Risk
Leveraging Engagement
Data to Predict Dropout
Psychological
Domain Educational
Exploring the Link Between Mental Domain
Health and Dropout Rates
Identifying At-Risk Students Through
Academic Performance
CONCLUSION
1
In conclusion, this project explored the potential of machine learning
algorithms to predict student dropout risk. By analyzing data
consisting of [list of features used (e.g., academic performance,
demographic information, engagement data)], the [STUDENT
DROPOUT PREDICTION BY USING MACHINE LEARNING]
achieved an accuracy of [86% - 92%] in identifying students at risk of
dropping out. This demonstrates the potential of machine learning as
a tool for early intervention and support in educational settings.
Thanks!
Overall, this project highlights the promise of machine learning for creating proactive strategies
to reduce student dropout rates. By leveraging these insights, educational institutions can
implement targeted interventions and support systems to help students succeed.
“Skills are sharpened, models take flight, pushing boundaries, day and night.”
CREDITS: This presentation template was created by Slidesgo, and includes
icons by Flaticon and infographics & images by Freepik
Please keep this slide for attribution
