Rapport
Rapport
1. Context
Student performance in exams is influenced by several factors beyond just study time. Family
environment, class attendance, access to educational resources, and motivation level all play a
crucial role. Understanding these factors helps identify key elements that promote academic success
and enables actions to be taken to improve student outcomes.
The data used in this project provides a comprehensive view of the elements affecting exam
performance. It includes information on study time, attendance, parental involvement, participation
in extracurricular activities, as well as other variables such as access to educational resources,
motivation levels, and the quality of teaching. Analyzing this data helps to better understand the
factors that can be leveraged to improve academic performance.
2. Main Objective
The main objective of this project is to build a predictive model capable of estimating a student's
final exam score based on a set of factors. This model will allow educators and institutions to better
understand the dynamics behind academic success and adopt strategies to optimize learning and
student performance.
3. Specific Objectives
• Identify the factors that have a significant impact on student performance.
• Predict the final exam score based on these factors.
• Provide recommendations based on the results to guide teachers, parents, and students
toward more effective practices.
4. Expected Results
• A high-performance predictive model: The project aims to develop several machine
learning models (linear regression, random forests, neural networks, etc.) capable of
accurately predicting students' final exam scores. The effectiveness of the model will be
measured in terms of accuracy and generalization ability.
• Identification of key factors: A detailed analysis of the data will identify the factors that
most influence academic performance, such as study hours, parental involvement,
participation in extracurricular activities, or access to educational resources.
• Practical recommendations: The model results will be used to provide concrete
recommendations for teachers and parents on actions to take to maximize student academic
performance. For example, if the model shows that parental involvement has a major
impact, strategies to strengthen that involvement could be proposed.
• Visualizations of variable relationships: Graphs and visualizations will be produced to
illustrate the relationships between different factors and student performance. This will
allow for easy interpretation of the results and effective communication of the conclusions.
5. About Dataset
The dataset used on this project provides comprehensive information on various factors influencing
students' academic performance. It includes both categorical and numerical data points, representing
a wide range of aspects related to student life, such as parental involvement, attendance, study
habits, sleep patterns, academic resources, and exam performance. The data is collected from
students and offers insights into how different personal, social, and environmental factors may
correlate with academic success.
5.1. Categoricals columns
Category/
Column Percentage Interpretation Conclusion
Range
Parental involvement is
Most students have
Parental moderate for most students,
Medium 50.89% medium parental
Involvement positively influencing their
involvement.
academic outcomes.
A smaller percentage High parental involvement
High 28.88% shows high parental can have a strong impact on
involvement. academic success.
A smaller proportion A small proportion may face
Low 20.24% has low parental challenges due to low
involvement. parental involvement.
Most students have
Access to Most students have sufficient
Medium 50.23% moderate access to
Resources resources for learning.
resources.
A significant number High access to resources
High 29.89% have high access to positively influences student
resources. outcomes.
A smaller proportion Limited access to resources
Low 19.87% has limited access to could hinder academic
resources. performance.
Majority of students
High participation could
Extracurricular participate in
Yes 59.60% enhance students’ personal
Activities extracurricular
growth and skills.
activities.
A significant
proportion does not
Lack of participation might
No 40.40% participate in
affect social development.
extracurricular
activities.
Most students have a Medium motivation could
Motivation
Medium 50.72% medium level of drive steady academic
Level
motivation. progress.
Low motivation may
A smaller percentage
Low 29.32% contribute to lower academic
has low motivation.
performance.
A smaller proportion of High motivation may lead to
High 19.96% students is highly exceptional academic
motivated. performance.
Internet Access Yes 92.45% Most students have Internet access is crucial for
internet access, academic success, and most
enhancing learning students have it.
Category/
Column Percentage Interpretation Conclusion
Range
opportunities.
A small percentage of Lack of internet access may
No 7.55% students lack internet hinder academic progress for
access. some students.
Almost half of the Income levels affect resource
Family Income Low 40.44% students come from access and overall academic
low-income families. opportunities.
A large portion of
Medium-income families
students come from
Medium 40.35% may have moderate access to
medium-income
resources.
families.
A smaller proportion of High-income students may
High 19.21% students come from have better access to
high-income families. educational resources.
Most students receive Adequate teaching quality
Teacher Quality Medium 60.12% medium-quality impacts most students
teaching. positively.
High-quality teaching could
A significant number
significantly improve
High 29.82% experience high-quality
students' academic
teaching.
performance.
A smaller proportion Low-quality teaching could
Low 10.06% experiences low- lead to poor academic
quality teaching. outcomes.
Public schools serve the
Most students attend majority of students,
School Type Public 69.59%
public schools. influencing the overall
educational environment.
Private schools may offer a
A smaller proportion
Private 30.41% different quality of education
attends private schools.
compared to public schools.
A large proportion is Positive peer influence
Peer Influence Positive 39.93% positively influenced fosters a good academic
by peers. environment.
A similar number
Neutral peer influence is
Neutral 39.23% experiences neutral
neither helpful nor harmful.
peer influence.
A smaller proportion is Negative peer influence
Negative 20.84% negatively influenced could harm academic
by peers. outcomes.
Most students do not Students without disabilities
Learning
No 89.48% have learning tend to have fewer barriers in
Disabilities
disabilities. learning.
A small proportion of A small portion of students
Yes 10.52% students have learning may require additional
disabilities. support.
Parental High School 49.46% Half of the students Parental education influences
Category/
Column Percentage Interpretation Conclusion
Range
have parents with a students' educational
Education Level
high school education. aspirations.
A significant number of
College-educated parents
students have parents
College 30.52% may provide more guidance
with a college
and support.
education.
A smaller proportion of Postgraduate-educated
students have parents parents may emphasize
Postgraduate 20.02%
with postgraduate higher educational
education. aspirations.
Most students live near Proximity to school can
Distance from
Near 59.39% their school, facilitating enhance regular attendance
Home
easier access. and reduce tardiness.
Many students live a
Moderate distances may still
Moderate 30.55% moderate distance from
allow for regular attendance.
school.
A smaller proportion
Long distances may pose
lives far from school,
Far 10.06% challenges to consistent
potentially affecting
attendance.
attendance.
Gender distribution is skewed
Gender Male 57.73% Most students are male.
slightly toward males.
A significant number of There is a notable proportion
Female 42.27%
students are female. of female students as well.
Mean/
Column Range Interpretation Conclusion
Median
Most students study
Students who study around
between 5 and 15 hours
10 hours weekly may
Mean: 10.5 per week, indicating a
Hours Studied 0 - 20 hours perform well academically,
hrs moderate amount of
but further efforts could
commitment to
help.
academics.
A smaller percentage Students who study more
20% of studies for over 20 hours than 20 hours may perform
> 20 hours
students per week, indicating high exceptionally, though
commitment. balance is important.
A small percentage do not
Lack of study time is a
5% of dedicate time to studying,
0 hours strong indicator of lower
students which could negatively
academic performance.
affect their performance.
Attendance 0 - 100% Median: The majority of students High attendance generally
85% attend classes regularly, leads to better academic
with a median of 85%. performance. Students
This suggests students should aim for 100%
Mean/
Column Range Interpretation Conclusion
Median
value regular class
attendance.
attendance.
A significant percentage Regular attendance is
70% of of students maintain an crucial for understanding
80 - 100%
students 80% to 100% attendance course material and
rate. participating in discussions.
A small percentage has
Low attendance can hinder
15% of poor attendance, which
< 80% learning and result in lower
students might negatively affect
grades.
performance.
3 - 10 hours Mean: 7 hrs Most students sleep
Adequate sleep contributes
between 6 and 8 hours,
positively to cognitive
Sleep Hours which is within the
performance and overall
recommended range for
health.
most adults.
< 3 hours A small number of
Insufficient sleep is
students sleep less than 3
2% of detrimental to academic
hours per night, which
students performance and overall
can affect health and
well-being.
academic performance.
> 10 hours Very few students sleep
Excessive sleep could be a
for more than 10 hours
3% of sign of poor time
per night, which could
students management or other
indicate issues with time
underlying issues.
management or lethargy.
0 - 100% Mean: 70% Most students score
A score of 70% is a solid
around 70%, with a large
performance, but aiming for
Exam Score variation depending on
higher results requires
study habits, resources,
consistent effort.
and participation.
15% of A smaller proportion of
students students achieves high High exam scores indicate
scores of 90% or more, mastery of the subject,
90 - 100%
indicating strong requiring dedicated study
academic ability and and understanding.
preparation.
10% of A small number of
students students score poorly, A score below 50% signals
indicating a lack of a need for intervention and
< 50%
preparation, low improvement in study
attendance, or other habits.
external factors.
1 - 12 Median: 5 Most students spend
Spending more time on
hours/day hrs between 3 and 8 hours on
campus may improve
Time Spent on campus per day,
engagement, but balancing
Campus balancing between
time is key for overall well-
classes, extracurricular
being.
activities, and study time.
Mean/
Column Range Interpretation Conclusion
Median
A small number of
students spend more than
Excessive time on campus
12 hours a day on
> 12 10% of might affect academic
campus, possibly due to
hours/day students performance or personal life
extracurricular
balance.
commitments or social
activities.
Most students prefer a
quiet and isolated A quiet environment
Study Quiet/ 55% of
environment for studying, supports concentration and
Environment Isolated students
which can improve focus better academic outcomes.
and productivity.
A moderate number of
students prefer group Group study can help some
Group 25% of study or a noisier students, but distractions
Study/Noisy students environment, which could should be minimized for
affect focus and others.
productivity.
A significant number of
students prefer the
The library is a conducive
20% of library, offering a more
Library environment for focused,
students structured and
quiet study.
distraction-free study
space.