Predicting the Impact of Sleep Patterns on Student Academic
Performance Using Machine Learning
Gayathri V P, Harisowndharya V, Haritha Saraswathy R and Gokul R
Dept. Information Technology, Kongu Engineering College (Anna University), Erode, India
Keywords: Machine Learning (ML), Decision Tree (DT), Support Vector Machine (SVM), K-Nearest Neighbor’s
(KNN), Random Forest (RF), Receiver Operating Characteristic (ROC), Area Under the Curve (AUC).
Abstract: Adequate sleep has a significant role in improving cognitive skills, especially memory retention. Sleep
deprivation at night and the consequent daytime sleepiness affect the physical and mental health of students
and their academic performance. The objective of this study is to build a predictive model using machine
learning and investigate the sleep pattern and its association with college students’ academic performance. It
is a cross-sectional study conducted among 375 undergraduate college students across various disciplines. A
questionnaire that contained questions on demography, sleep habits, academic performance, and ideal sleep
was used to collect data. The supervised classification algorithms such as Decision Tree, Support Vector
Machine, K-Nearest Neighbors, Naive Bayes and Random Forest are used to classify and predict the effect
of sleeping behaviours on college students’ academic performance. It is determined that the Decision Tree
prediction model has an accuracy of 97.33% in the classification prediction of academic performance. It is
observed that sleep duration is positively correlated with better academic performance. Key predictive factors
for academic performance include sleep hours, sleep quality, stress levels, and electronic device usage before
bed. These findings provide quantitative, objective evidence that better quality, longer duration, and greater
sleep consistency are strongly associated with better academic performance in college. The Decision Tree
model proves to be highly effective, emphasizing the relevance of sleep habits in predicting academic
outcomes. This implies that educators can use certain sleeping habits to improve student support services and
increase the overall effectiveness of the educational system. Our work involving research on sleep patterns
and the direction of sustainable development in academic performance training proved to be encouraging.
1 INTRODUCTION
1.1 Background
Sleep is essential to well-being and the brain as it
improves memory, learning processes and anger
mitigation. That is why many college students
experience some problems with getting a proper
amount of sleep because of studying and other
activities, and the bad quality sleep causes such
consequences as worsened ability to pay attention and
solve problems, decreased academic achievement, as
well as increased stress and anxiety. This action
research-based study employs a machine learning-
based approach that seeks to make predictions
regarding student sleep patterns and their impact on
students’ academic performance to improve student
sleep and advancement.
1.2 Problem Statement
Although it is recommended that people, especially
students, should have at least 7-8 hours of sleep every
night for proper learning and growth of the brain,
most college students lack adequate sleep due to the
pressure arising from volumes of assignments and
other leadership tasks, use of electronic devices at
night. Inadequate amounts of sleep deprive active
learning and memory, critical thinking, and therefore
cause lower performance at school. Most past surveys
have elicited responses concerning the number of
hours the participant sleeps and failed to consider
elements such as the quality of sleep, the amount of
time taken to get to sleep, and devices. Stress and
behavioural factors also influence the interaction
between sleep and academic performance besides
behavioural factors. This study seeks to address these
gaps by modelling and assessing various sleep factors
on students’ performance using the ML technique.
862
V P, G., V, H., R, H. S. and R, G.
Predicting the Impact of Sleep Patterns on Student Academic Performance Using Machine Learning.
DOI: 10.5220/0013606200004664
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 862-870
ISBN: 978-989-758-763-4
Proceedings Copyright © 2025 by SCITEPRESS – Science and Technology Publications, Lda.

They aim to find out relevant variables that affect
sleep and how they can recommend strategies for
improved sleep and performance.
1.3 Research Objectives
Objective 1: Analyzing the existing data on the
regulation of sleep cycle patterns, consider the
application of the machine learning technique to
forecast trends of such patterns in correlation to
students’ performance.
Objective 2: Determine which aspects of sleep
including duration of sleep, sleep quality, timing, and
pre-sleep activities correlate with academic
performance at night most strongly.
Objective 3: This shall entail the use of a number
of machine learning algorithms on the sleep patterns
data and then evaluate their performance to determine
which one of them will give the best for predicting
the performance in academics. In that regard, the
current study will help to fill gaps in knowledge on
sleep and machine learning.
1.4 Significance of The Study
This research is important because there is little
research on the findings on hours of sleep and
academic performance of college students.
Discussing the quality, quantity, and rhythm of sleep
further, the nature of its effects on cognitive processes
essential for academic performance is also described.
The research can be applied to legislative measures
by schools to promote adequate sleep knowledge and
practice as well as suitable academic schedules to
improve the learner’s achievement and well-being.
Besides, it employs machine learning to identify
certain actions related to learner outcomes that can
help in developing early support mechanisms for low
achievers. In summing up, this study fills some of the
gaps in the literature with reference to theories and
proffers solutions that can be implemented by
educators, counsellors, and parents to enhance learner
performance.
2 LITERATURE REVIEW
2.1 Sleep and Academic Performance
Relationship
Public Health pointed out that chronic sleep
deprivation affects the student’s learning capacity as
well as their physical and mental health thus seeing
their poor sleep pattern as being attributable to the
poor performance. To this effect, later school start
time was also reported to positive impact on sleep
duration and quality and hence the improvement of
academic performance (Alfonsi, Scarpelli, et al. ,
2020). In another study conducted equally in Sleep
Advances, the authors provided similar evidence
covering the fact that learners with poor sleep regimes
have fewer academic performance outputs than their
counterparts who wake up refreshed. The review also
establishes that quality of sleep as opposed to the
number of hours spent sleeping contributes to
academic success. The lack of sleep disorders as well
as sleep consistency and sleep hygiene were found to
be strong predictors of better academic performance
(Falloon, Bhoopatkar, et al. , 2022).
As a result, schools and governments should
provide programs for improving the effectiveness of
students’ sleep. Interventions consist of increasing
consciousness about sleep, encouraging sleep, or
shifting the time and school hours according to
students’ strophe period. The use of such approaches
can assist in unleashing the academic capacities of
students and also foster their whole-sided growth.
Thus, this research will build on previous work by
determining which aspects of sleep are the most
relevant to performance, as well as what aspects
should be targeted by educators and policymakers.
2.2 Machine Learning in the Context of
Education
It has become almost impossible for research on
education to be conducted without the use of machine
learning as it offers the most effective way of
evaluating students’ performance besides helping in
the identification of any vulnerable persons. Such
models deal with large data and can capture
dependencies other than via statistical measures
conventional for classical statistics.
2.2.1 Overview of Machine Learning
Applications in Education
Technology is being used primarily for prediction and
for identifying a student who requires intervention
and they are effective in analyzing educational data.
Basic supervised learning algorithms such as
Decision Tree, Naive Bayes and Random Forest have
been found to be useful for this. In the experiment,
the Decision Tree model is the best model that
provides an accuracy of 97.33% and is better than
Naive Bayes (92.00%) and Random Forest (92.00%)
in student classification by the academic performance
Predicting the Impact of Sleep Patterns on Student Academic Performance Using Machine Learning
863

criteria. A study by Aggarwal et al. (2019) pointed out
that these predictive models could successfully
classify the students so that the teacher could
intervene on time. Furthermore, Hasan et al. (2021)
proved that artificial intelligence methods could
improve individualisation in learning. In conclusion,
the result shows that the decision Tree model gives
the best prediction of academic performance among
all the four models.
2.2.2 Recent Research and Development
Thus, the observed present trend in using machine
learning increases in educational environments is
backed by recent studies and observations. For
instance, Hasan et al. (2021) described how the
pedagogical mobile applications that apply machine
learning and more specifically deep learning
algorithms, were capable of predicting the
performance of students with relatively high levels of
accuracy(Webb, Fluck, et al. , 2021). This capacity to
manipulate massive-size educational data and find
out more profound patterns is especially important as
it enables us to get more meaningful characteristics of
student learning behaviours. But the use of deep
learning in education is still in its infancy and it has
some problems, examples of which are the
requirement of large labelled datasets, and the
problems associated with the explanation of these
models. Zaffar and collaborators, for instance, talked
about the performance of feature selection methods to
support EDM, illustrating that feature selection is
critical to the performance of a model(Webb, Fluck,
et al. , 2021).
2.2.3 Impact on Educational Practices
The process of integration of machine learning into
educational practices is revolutionary. In this way,
with the help of researchers’ data, educators can
identify the student’s needs and intervene more
effectively in the situation to enhance educational
results. Nevertheless, problems like data privacy and
possible biases in the work of an algorithm are still
essential subjects when it comes to the further
development of these technologies (Zhai, Lu, et al. ,
2023), (Kakkad, Shingadiya, et al. , 2023). Summing
up it can be pointed out that machine learning
Figure 1: Flowchart
enables redefining educational processes as the rich
set of tools assisting in the analysis of student
outcomes. With the advancement of this technology,
it is foreseen that it will have a broader implication in
education though with its implication comes both
opportunities and challenges that will be encountered
by both the educationist and researcher.
2.3 Existing Gaps of Research
Where a considerable amount of research has helped
in the demonstration of the relationship between sleep
and academic performance, still very few studies
have applied machine learning to forecast academic
success based on sleep behaviour.
INCOFT 2025 - International Conference on Futuristic Technology
864

2.3.1 Absence of Predictive Models
Most of the focused studies report the relationship
between the patterns of sleep and the grades without
using machine learning for the final prediction. For
example, Okano et al. found positive impacts of time
spent asleep on academic success but did not apply a
predictive modelling technique (Okano, Kaczmarzyk,
et al. , 2019). Likewise, Chen et al. established the
effect of sleep deprivation on the associated grades at
school but did not develop predictive models using
machine learning (Jalali, Khazaei, et al. , 2020).
2.3.2 Challenges in Integrating Data
Despite there being elaborate data on sleep, the
integration into the machine learning models yet falls
short. Zhou, et al (2021), made attempts at
educational predictions through machine learning, but
still, they were not able to completely integrate
sleeping behaviours in these models. This proves a
challenge towards sufficient size and quality in data.
The majority of studies have fallen back on a narrow
set of machine-learning algorithms. For example,
Trockel et al. (2020) have applied Random Forests to
predict academic performance, but comparison with
other algorithms like Support Vector Machines or
Neural Networks has not been made.
2.3.3 Underutilization of Advanced
Techniques
Advanced machine learning techniques are not
exploited at all in this particular domain. On one
hand, we can see the various research areas' impact on
deep learning in Lee et al. (2021), but they did not
apply it to educational data and sleep patterns
(Hernández, Antonio, et al. , 2019). This study seeks
to fill these gaps by applying different machine
learning algorithms for the prediction of academic
performance using fine-grained sleep data to better
understand and identify effective predictive models.
3 METHODOLOGY
3.1 Study Design
The study used a cross-sectional research technique
whereby data were collected from participants at one
time to investigate the relationship or correlation
between sleep patterns and academic performance
among undergraduate students. To increase external
validity the data were collected from students of
different faculties and in different academic years.
Cross-sectional designs are very efficient in
identifying relationships between variables like; sleep
behaviours and academic performance, without
necessarily extending their study over long periods.
Unfortunately, they do not justify cause-and-effect
relationships. Further studies, employing longitudinal
research designs are advised, to give a causal account
of the relationship between sleep patterns and
performance. Concisely, the present study examined
the effects of sleep patterns on student’s performance
providing useful information for instructors and sleep
management for learners within a given population
group.
3.2 Participants
A cross-sectional research design was adopted, using
375 participants comprising undergraduate students
across the faculties of art, science and the social
sciences. Participants were recruited based on their
age 18-24 which is currently the typical students’ age
struggling with the challenge of college education.
This was with the intention of comparing sleep
patterns and academic performance between male
and female students to increase the generality of the
findings to the entire college populace. Hence, the
sampling technique used in this study was purposive
and convenient in that easily accessible students were
targeted in addition to other categories. Each
participant reported their consent to take part in the
study, in adherence to the ethical requirement.
3.3 Data Collection
Data collection for this study utilized a structured
questionnaire designed to capture a wide range of
factors related to sleep and academic performance.
The questionnaire comprised several sections is
shown in Table 1:
1. Demographics: Collected data on
participants’ age, gender, and field of study
to ensure diverse representation.
2. Sleep Patterns: Included questions about
sleep duration, quality, bedtime, wake time,
and nap hours, focusing on participants' sleep
routines.
3. Sleep Quality: Assessed through a visual
analogue scale and inquiries about
difficulties falling asleep, waking during the
night, and the use of sleep aids.
4. Academic Performance: Evaluated
participants’ self-reported academic success,
Predicting the Impact of Sleep Patterns on Student Academic Performance Using Machine Learning
865

5. GPA, and frequency of sleepiness in class,
including whether they skipped classes due to
sleep issues.
6. Additional Factors: Explored variables such
as caffeine intake, exposure to screens,
perceived academic stress, and health
conditions affecting sleep.
Table 1: Participant Data
Participant Characteristics n (%)
Age
18-20 100 (26.7%)
21-23 150 (40.0%)
24-26 90 (24.0%)
27 or older 35 (9.3%)
Gender
Male 150 (40.0%)
Female 225 (60.0%)
Field of Study
Engineering 120 (32.0%)
Business 100 (26.7%)
Science 80 (21.3%)
Arts 75 (20.0%)
Average Sleep Hours
Less than 6 hours 50 (13.3%)
6-7 hours 200 (53.3%)
7-8 hours 100 (26.7%)
More than 8 hours 25 (6.7%)
Quality of Sleep
Good 150 (40.0%)
Average 130 (34.7%)
Poor 95 (25.3%)
Academic Performance
High 75 (20.0%)
Medium 200 (53.3%)
Low 100 (26.7%)
This comprehensive approach ensured the
collection of relevant variables, crucial for developing
prognostic and diagnostic models, and enhancing the
study's credibility for practical educational
applications.
3.4 Machine Learning Models
In this research, five classifiers – Decision Tree,
Support Vector Machine (SVM), K-Nearest
Neighbors (KNN), Naïve Bayes, and Random Forest
Classifier – were employed to forecast academic
performance using sleep data. These algorithms were
considered optimal for educational prediction tasks as
well as for handling data distributions and relations.
The models were then compared with a view of
determining the best sleep profile predictor of
academic performance. DT came out as the most
performing, revealing better performance to the other
models, indicating its capability to capture vital
characteristics in educational data patterns.
Table 2: Best Qualities of the Model
Model Best Qualities
Decision Tree
Interpretability: Provides clear and
easy-to- understand decision rules
and visualizations.
Handling Non-linear Data:
Effectively models complex non-
linear relationships between
variables.
-Feature Importance: Automatically
identifies and ranks the most
important features for
p
rediction.
Support Vector
Machine (SVM)
Effective in High Dimensions:
Performs well with a large number
of features.
Robust to Outliers: The margin
maximization approach handles
outliers effectively.
Kernel Trick: Capable of non-linear
classification usin
g
the kernel trick.
K-Nearest Neighbors
(KNN)
Simplicity: Easy to understand and
implement.
No Training Phase: KNN is a lazy
learner, meaning it requires no
training phase.
Flexible: Can be used for both
classification and
regression.
Naive Bayes
Speed: Extremely fast in both
training and prediction phases.
Handles Missing Data: Can handle
missing data and noisy data well.
Random Forest
High Accuracy: Often achieves high
accuracy due to ensemble learning.
Robustness: Resistant to overfitting,
especially with a large number of
trees.
Feature Importance: Provides
insights into
feature im
p
ortance.
3.5 Model Evaluation Metrics
To evaluate the predictive power of the machine
learning models, several key metrics were used:
1. Accuracy: Estimate the quantity of right
classifications but can be distorted in imbalanced
sample space.
2. Precision: The number of actual positive cases
INCOFT 2025 - International Conference on Futuristic Technology
866

identified to the number of positive cases as
estimated by the model. When the cost of false
positives is high, then high precision is desirable.
3. Recall: Frequently called sensitivity, it expresses
true positives divided by all actual positives and
shows how effective the model is at identifying
positives. The high recall is important when false
negatives are severely expensive.
4. F1 Score: Standard deviation of precision and
recall, helpful for analyzing the difference
between precision and recall, usefull in the
situation when one type of error predominates and
the goal is to minimize it.
5. ROC Curve and AUC: ROC analysis is used for
the evaluation and comparison of the binary
classification model and AUC demonstrates the
overall performance of the system with higher
AUC means beteer classification of the class.
In order to archive the highest accuracy in model
performance, an algorithm called
RandomizedSearchCV was used because it randomly
selects hyper parameters for evaluation and provides
a clear correlation between sleep/awake cycles and
academic performance.
Table 3: Performance Metrics Formula
Metric Formula
Accuracy Accuracy = (TP+TN) /
(
TP+FP+FN+TN
)
F1 Score F1 Score = 2×Precision +
Recall / Precision × Recall
Precision Precision = TP+FP / TP
Recall Recall = TP+FN
/
T
P
ROC Curve and
AUC
AUC=

𝟏
𝟎
𝑅𝑂𝐶 Curved (False
Positive Rate)
4 RESULTS
4.1 Descriptive Statistics
This research showed students had poor sleeping
habits many of whom went to bed without getting 7-
8 hours of sleep a night which is bad for cognitive
function and academic performance. More
specifically, 55% that they get not more than 7 hours,
and 20% above that which seems erratic and has a
negative impact on self-reported performance.
However, the students with high sleep efficiency and
stable timetables achieved higher results; the
students, who slept 7-8 hours, top graded 30% against
15% of the students, who slept no more than 5(6)
hours.
Further, in regard to nutrition and exercise, 34% of
them mentioned having bad nocturnal routines. Out
of users suffering from sleep problems, or using
devices late into the night, 68 per cent reported always
feeling fatigued in class and would therefore not be
able to concentrate well. This goes a long way to show
that people could be very ruined if they do not
practice good sleep hygiene if they are to get good
grades in their academics. These results provide the
basis for additional analysis in view of machine
learning models.
Figure 3. Correlation Heat map for overall Features.
4.2 Major Indicators of Performance
The analysis identified several key factors
influencing student's academic performance is shown
Figure 3:
1. Sleep Duration: Those attaining at least 7-8 hours
of sleep recorded a positive performance while
students who hours of sleep recorded a negative
performance.
2. Sleep Quality: Sustained sleep with little
interference was positively associated with
students’ performance and poor quality of sleep
affected their concentration and memory.
3. Stress Levels: More stress amplifies the
disturbance in performance and sleep that are
mutually consequential. These effects can,
however, be reduced through good stress
management.
4. Electronic Device Usage Before Bed: Pre
Predicting the Impact of Sleep Patterns on Student Academic Performance Using Machine Learning
867

bedtime device use meant more awakenings and
so more detrimental classroom outcomes due to
disruption from blue light that prevents
melatonin secretion.
In conclusion, the study focuses on the need for
sufficient sleep, good quality sleep, reduction of
stress levels, and avoiding the use of devices before
going to sleep to enhance academic performance;
gives working recommendations for students teachers
and instructors.
5 DICUSSION
5.1 Interpretation of Results
Therefore, the studies presented in the paper stress
the significance of quantity and qualitative sleep for
college students’ performance. The Decision Tree
model had an accuracy of 97.33% as opposed to all
the other models and succeeded in capturing the
interactions among the sleep factors, duration,
quality, stress level, electronic device usage, and the
AUC of 0.9830. These outcomes support related
research indicating that sleep regularity and length
tend to improve memory and focus – crucial for
learning. Also, any use of electronic equipment at
night has a very negative impact of the quality of
sleep.
In conclusion, let it be pointed out that the present
work focused on the impact of sleep on learning and
on the applicability of the Decision Tree when it
comes to the prediction of educational performance.
The implication is that better sleep quality and sleep
patterns will lead to much better academic
performance.
5.2 Comparison with Existing Literature
The results of the present study support a great
amount of prior literature that sheds light on the need
for sleep and its effects on cognition and academic
achievement. A great deal of research has pointed out
the fact that sleep is important in memory
consolidation and learning. For example, an article by
Hershner and Chervin (2020) explained that sleep
deprivation impairs attention, memory, and emotional
functioning, which are important factors students
require to excel in class. In the same manner, Lo et al.
(2021) showed that sleep quality is related to
academic performance, most especially in courses
that demand more engagement of the brain. However,
the present research contributes to this line of
research by developing a model that may help in the
early identification of students likely to perform
poorly in their studies because of faulty sleep
patterns. Therefore, distinguishing from prior works
that reveal associations between sleep and academic
performance, this study aims at developing a machine
learning model to predict such performance, and,
hence, approach the problem more proactively.
For example, the Decision Tree algorithm
employed in this study created a sound explanatory
model with an overall accuracy of 97.33% and AUC
of 0.9830. This methodological approach is similar to
other research in educational data mining, for instance,
the work by Yousef et al. where they used comparable
machine learning algorithms to predict student
success from their behaviors. However, this research
is novel in comparing the overall sleep of students to
its components including the duration, quality of
sleep, stress levels, and the use of electronic devices
before going to sleep and how they affect the
performance of students. The same has been also
pointed out by Wang, et al. (2021) who stated that an
irregular sleep schedule was more deleterious to
cognitive functioning in keeping with the present
research.
In sum, the findings confirm the prior work on
sleep importance and build upon it by offering the
prognostic approach that might be utilized to deliver
specific modifications. This advancement could
prove to be promising in improving educational
results since it can work to handle sleep-related issues
preventatively.
Figure 4. Importance for Academic Performance.
5.3 Applications to Educators and
Policy makers
This research reveals the necessity of the return to
excellent performance schools for additional
consideration for sleep education. These should
include sleep hygiene education that engages students
in programs such as sleep schedules, enough sleep
INCOFT 2025 - International Conference on Futuristic Technology
868

and no screen time before bedtime, sleep workshops
and changing the school start time. This paper
recommends that policymakers should use
educational materials to augment sleep yet pursue
academic excellence introducing health-friendly
dormitories and funding more research on sleep and
learning. Improving sleep quality indeed leads to
better academic performance and better quality health
making schools a place where students blossom
academically as well as psychologically.
6 CONCLUSION
6.1 Summary of Findings
Drawing from this paper, it becomes evident that
college students with regular sleep, adequate sleep
and quality sleep have better academic results as
compared to their counterparts. In terms of accuracy
and AUC, the Decision Tree demonstrated a better
result compared to SVM, Random Forest, KNN, and
Naïve Bayes to identify the relationship between
chronological sleep features (duration, quality, stress
levels and electronic devices) and performance.
These data buttress the need to pay adequate attention
to the amount of sleep one takes when learning or
performing tasks. The study’s models may be used by
educators and policymakers to know the students who
usually have inadequate rest and who may require
specific remedial actions to overcome their struggling
academic performances and poor general health.
6.2 Recommendations
From previous scholarly research, educational
interventions presented to improve students’
performance should aim at helping students get better
quality sleep. Decision makers need to encourage
improved sleep habits, as proper sleep improves the
brain’s capability to perform some functions such as
memory. Stress has to be lessened also, as it interferes
with the quality of sleep. These measures may
comprise employee services such as fitness and
health promotion activities and counselling services;
stress management strategies including relaxation
and meditation. Moreover, restrictive use of
electronic devices at night is crucial because blue light
hampers the production of melatonin. Education
curricula should alert students of these effects, and
teach them how to practice, for example,
minimization of screen time. It is believed that the
result of these recommendations will enhance
students’ performance.
6.3 Limitations of the Study
A small limitation is that data is obtained from self-
completion of questionnaires, which can result in
reporting biases such as recalling their improved
sleep, stress and academic performance. Also, the
cross- sectional design method limits causality and
temporal variations as the data are collected in a
single point without controlling for change
within weeks or semesters. It is recommended that
future studies should incorporate longitudinal designs
to monitor changes in sleep behaviors and their effects
on performance. These may reduce reliability due to
bias that results from self- reporting, but integrating
objective data from wearables may reduce such
weaknesses.
6.4 Future Research Directions
Future studies should also determine the impact of
intrusive sleep interventions with learners especially
in the academic arena. Whereas this study establishes
a relationship between sleep and performance, future
experimental studies should help understand the
effects of sleep knowledge, stress reduction and
technological devices. The more complex machine
learning algorithm, for instance, deep learning
algorithms or an ensemble of a variety of algorithms
could further augment the analysis of sleep
behaviours. Further, the use of tending data from
wearables may enhance the prediction quality of the
algorithm. The results would be more generalizable if
the study recurs in people of various ethnicities, ages
and learning environments. They may result in
positive approaches towards controlling student sleep
as well as improving learning outcomes.
REFERENCES
Alfonsi, V.; Scarpelli, S.; D’Atri, A.; Stella, G.; De
Gennaro, L. Later School Start Time: The Impact of
Sleep on Academic Performance and Health in the
Adolescent Population. Int. J. Environ. Res. Public
Health 2020, 17, 2574.
Karen Falloon, Harsh Bhoopatkar, Fiona Moir, Miriam
Nakatsuji, Andy Wearn, Sleep well to perform well: the
association between sleep quality and medical student
performance in a high-stakes clinical assessment,
SLEEP Advances, Volume 3, Issue 1, 2022, zpac019.
Zhai X and Lu M (2023) Editorial: Machine learning
applications in educational studies. Front. Educ.
8:1225802. doi: 10.3389/feduc.2023.1225802
Anju Kakkad, Chetan Shingadiya; A role of machine
learning algorithm in educational data mining using
Predicting the Impact of Sleep Patterns on Student Academic Performance Using Machine Learning
869

predictive models: A review. AIP Conf. Proc. 14
November 2023; 2963 (1): 020013.
Webb, M.E., Fluck, A., Magenheim, J. et al. Machine
learning for human learners: opportunities, issues,
tensions and threats. Education Tech Research Dev 69,
2109–2130 (2021).
Okano, K., Kaczmarzyk, J.R., Dave, N. et al. Sleep quality,
duration, and consistency are associated with better
academic performance in college students. npj Sci.
Learn. 4, 16 (2019).
Jalali R, Khazaei H, Paveh BK, Hayrani Z, Menati. L. The
Effect of Sleep Quality on Students' Academic
Achievement. Adv Med Educ Pract.
Ojajuni, Opeyemi & Ayeni, Foluso & Akodu, Olagunju &
Ekanoye, Femi & Adewole, Samson & Ayo, Timothy
& Misra, Sanjay & Mbarika, Victor. (2021). Predicting
Student Academic Performance Using Machine
Learning.
Trockel MT, Barnes MD, Egget DL. Health-related
variables and academic performance among first- year
college students: implications for sleep and other
behaviors. J Am Coll Health.
Hernández, Antonio & Herrera-Flores, Boris & Tomás,
David & Navarro-Colorado, Borja. (2019). A
Systematic Review of Deep Learning Approaches to
Educational Data Mining. Complexity.
INCOFT 2025 - International Conference on Futuristic Technology
870