Literature Survey on Comparative Analysis of Machine Learning Techniques

Abstract

In the educational field predicting student performance importance and can provide results that can assist educators to plan accordingly. When students who are at risk of not performing well are identified beforehand strategies can be planned accordingly. For this purpose student performance prediction is very useful however, manual assessment of previous grades is a tedious and time consuming task. We have presented a detailed study where ML techniques are implemented to predict the student academic performance. Many supervised machine learning algorithms that are tree based, probabilistic, distance based and margin based and many others that are implemented are studied. Techniques like feature selection and preprocessing reduce noise and bias. The paper also explains how accurate performance prediction can help educational institutions in planning academics and guiding students that need support.

Introduction

The text discusses the use of machine learning (ML) to improve student performance prediction in modern digital education systems. With the growth of online learning platforms and academic systems, large amounts of student data (such as grades, attendance, behavior, and demographics) are generated, making manual analysis inefficient and error-prone.

Traditional methods, which rely mainly on exam scores, fail to capture the complex factors influencing performance, such as learning habits and socio-economic background. ML techniques address this by analyzing large, complex datasets to identify patterns and predict outcomes more accurately, enabling early identification of underperforming students and timely interventions.

Various ML approaches are used:

• Supervised learning (e.g., Decision Trees, SVM, Random Forest, Naive Bayes) for classification tasks like pass/fail prediction

• Unsupervised learning (e.g., K-means, ANN) for discovering hidden patterns

• Deep learning (e.g., CNN, RNN) for advanced pattern recognition

• Dimensionality reduction (e.g., PCA) to improve efficiency

Research shows that ML models outperform traditional statistical methods, especially when combined with proper preprocessing (handling missing data, normalization, encoding). Ensemble methods like Random Forest improve accuracy and reduce overfitting, while explainable AI techniques (e.g., SHAP, LIME) enhance transparency in decision-making.

The study highlights challenges such as data complexity, imbalance, privacy concerns, and the need for scalable systems. However, ML-based predictive analytics can significantly improve educational outcomes by enabling personalized learning, reducing dropout rates, and supporting institutional planning.

The proposed methodology involves applying multiple ML algorithms (e.g., Linear Regression, SVM, XGBoost) to historical student data, evaluating their performance, and selecting the best model. Cross-validation ensures reliability and scalability across different academic environments.

Conclusion

Data related to education and digital learning platforms is available in abundance .Academic records are present in many forms. It is an inefficient and tedious job to manually analyze this data and may cause errors too. A solution to this problem can be ML techniques as they utilize data mining for the purpose of prediction. Accurate prediction is essential for educators to identify the students who are probable to underperform and design a plan to implement corrective measures. Emotional and psychological factors also influence a stu-dent’s performance like socio-economic background, learning habits and past performance which too needs to be analyzed. Machine Learning algorithms are designed in a manner such that they handle these relationships too. Predictive models are trained on historical data, meaningful patterns are extracted that help in decision-making process.ML techniques help to improve accuracy. Digital transformation has changed the education field and has led to generation of tremendous amount of data.ML techniques are capable of handling this large amount of data and the variations present within this data. Models are trained on historical student data, meaningful patterns are discovered that support to make decisions. The integration of ML techniques also helps to improve the pre-diction accuracy, reduce bias, and design data driven educational strategies. In recent years, there has been a rapid digital transformation in the educational field due to the adoption of learning platforms in online mode,LMS and academic sys-tems. These platforms continuously generate large volumes of structured and unstructured data related to students’ academic performance, attendance, learning behavior, assessment results and interaction patterns. The extraction of meaningful insights from this data is a challenge for institutions that aim to boost learning outcomes reflected through student success rates. The traditional methods may rely only on exam scores and such methods do provide a basic understanding of student’s academic status yet they may not capture the complex fac-tors that influence their academic performance. Factors like attendance consistency, learning habits and socio economic background play a significant role which may be missed. Moreover, student performance prediction is important to improve the learning effectiveness and overall education qual-ity in educational institutions. The effectiveness of teaching learning can be improved when gaps and identified and remedies are designed to them on time. Educational institutions are now using intelligent decision support systems that help faculties to monitor student performance at an individual level and design customized teaching strategies for them. Scalability is required in institutions where analysis of data of thousands of students is simultaneously done.Machine learning based systems learn from these patterns and can handle diverse curriculum structures across different institutions. Fig. 1 illustrates the of machine learning techniques considered in various studies that are organized into major cat-egories. There exist four machine learning techniques namely supervised learning, unsupervised learning and deep learning and dimensionality reduction methods. Supervised learning methods depend on labeled data to predict outcomes they include Na¨?ve Bayes (NB),Random Forest (RF),k-Nearest Neighbors (KNN) and Support Vector Machines (SVM). Unsupervised learning methods are used to find hidden Fig. 1. Machine Learning Techniques structures in unlabeled data they include Artificial Neural Networks (ANN), Fuzzy C-Means (FCM) and K-Means Clus-tering (KMC). Deep learning is a subset of machine learning which in-cludes Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) which are useful to recognize pat-terns. Principal Component Analysis (PCA) and Adaptive Non-linear Mapping Methods (ANMM) are also implemented in various studies. This classification gives a complete overview of the algorithms implemented and studied. II. LITERATURE REVIEW Student performance prediction has become very important in recent times due to the growth of digital academic systems and learning platforms [8].Educational data includes academic scores, attendance behavioral patterns and demographic infor-mation [6].Accurate prediction is a difficult job as multiple factors that influence a student’s performance are present [10].Traditional methods that are based on manual evaluation and statistical analysis are limited and cannot handle complex datasets [15].These approaches fail to identify students who are at risk [11].Researchers have explored automated predic-tion techniques using data-driven models [13].Machine learn-ing techniques enables efficient analysis of large educational data [1].These techniques improve the prediction accuracy and help in early academic intervention [14]. Classification-based models are commonly preferred for predicting categorical outcomes such as pass or fail [9].De-cision tree algorithms are widely used because of their simple structure and interpretability [7].They help to identify key attributes that influence student performance, such as atten-dance and internal assessments [2]. Their transparent decision-making process makes them suitable for educational applications [6].Probabilistic models like Naive Bayes were studied because of their simplicity and low computational cost [3]. Despite assuming feature independence, these models show competitive results on preprocessed datasets [5]. They are par-ticularly useful for baseline performance prediction [12].Sup-port Vector Machine models were introduced to handle non-linear patterns in student data [6].Research showed that the performance of SVM performs is good in feature spaces of high dimension [3].Proper kernel selection also improves the classification performance [10].Due to increase in data complexity ensemble learning techniques gained importance in educational data mining [1].Random Forest models improve the accuracy as they combine many decision trees together for analysis [2].These models reduce the overfitting [9].Data preprocessing plays a crucial role in student performance prediction [5]. Techniques like normalization, handling miss-ing value and encoding significantly affects model outcomes [4].Poor preprocessing may cause bias in results and may re-duce the prediction reliability [15].Recent studies highlighted the role of predictive analytics to reduce the dropout rates and improve academic planning [6].Early identification of weak students helps institutions to provide academic support [7]. Such approaches improve overall outcomes of the teaching learning process [11]. Large scale of academic and behavioral data is generated due to rapid digitalization of educational institutions which makes student performance prediction a research problem to work on [8].Datasets that are diverse and massive in nature are generated through modern academic systems that contain data related to students and their academic activities which is very precious to predict future outcomes [6].However,owing to the multidimensional nature of educational data, it is difficult to accurately predict student performance [10]. Early studies on this topic mainly focused on traditon-ally used statistical techniques and manual evaluation which provided limited capability to predict outcome [15].These approaches were unable to capture the relationships among academics,behaviour and other factors [11].Therefore,such tra-ditional methods mostly failed to accurately predict perfor-mance of students and did not provide a clear result to design a strategy [7]. Thus to find better solutions,researchers began to adopt machine learning techniques to perform data-driven student performance prediction [13]. ML models can be trained on educational data collected over a period of time and do not need any extra formulation and they can yet extract patterns from it [1].Several studies have demonstrated that ML techniques are better than statisti-cal methods used traditionally in terms of prediction accuracy [14]. Classification-based machine learning models are widely used for prediction of student academic prediction,as academic outcomes are represented as categorical labels such as pass, fail or grade levels [9].These models provide insights that can be used to design academic planning, counseling and personalized learning strategies [6]. Comparative studies have shown that classification techniques are more effective than regression-based models when discrete performance categories are present [10]. Decision Tree algorithms have been extensively studied as they are interpretable and are able to identify key attributes that affect student performance [7].These models generate hierarchical decision rules that help educators to understand how factors such as attendance, internal assessments and prior grades influence academic outcomes [2].The transparent struc-ture of decision trees makes them particularly suitable for educational environments where things need to be explained [6]. Probabilistic models such as Naive Bayes have also been applied to student performance prediction because of their simplicity and low computational requirements [3].Despite the assumption of conditional independence among features here,Naive Bayes classifiers have achieved good results on well-preprocessed educational datasets [5].These models are often implemented as baseline classifiers for performance comparison in predictive studies [12]. Support Vector Machines have been introduced to handle decision boundaries that are not linear and feature spaces dimensionally high in educational data [6].Research indicates that SVM classifiers perform effectively when appropriate kernel functions are selected and hyper-parameters are used in an optimized manner [3].Kernel-based learning helps SVMs to model relationships between student attributes and perfor-mance outcomes [10]. Educational datasets are growing in size and complex-ity,ensemble learning techniques have gained significant at-tention in recent studies [1].Random Forest models combine many decision trees for improving the prediction accuracy and reduce variance in the model [2].Empirical evaluations have shown that ensemble-based approaches generally outperform individual classifiers as they improve generalization [9]. Several studies have found the that preprocessing data is important for increasing the effectiveness of prediction models [5].Preprocessing techniques such as normalization, handling missing values and encoding attributes of categorical nature significantly impacts the model performance [4].Inadequate preprocessing may lead to bias and noise which furthur may lead to misleading conclusions [15]. Recent research describes the role of predictive analytics to reduce dropout rates and improving educational outcome [6]. Early identification of low-performing students helps institutions to provide aca-demic support [7]. However, existing literature also points out to challenges like data imbalance,privacy concerns and lack of standardized evaluation frameworks, thus more research is required in this domain [13]. Recent studies have also shown a growing interest in applying advanced ML methods for performance prediction in an more accurate manner in higher education environments [17].There is a huge availability of large and diverse educa-tional datasets for researchers to study different models that are capable of identifying learning patterns that traditional methods may miss out [16].These modern approaches improve both, the prediction accuracy as well as usability in decision-making systems [20]. Several researchers have emphasized that deep learning models can relationships between academic and behavioral features [16].They learn hierarchical representations from student data and show good performance as compared to traditional machine learning algorithms [19]. However, their effectiveness largely depends on data quality,model tuning and availability of sufficient training samples [17]. The integration of predictive analytics into systems is a key application of prediction of student performance research [18].Such systems continuously monitor student progress and thus institutions can intervene before academic difficulties become severe [18]. If students who are at risk are identified early,it improves the retention rate and academic success [17]. Machine learning approaches implemented in hybrid man-ner have overcome the limitations of individual classifiers in educational prediction tasks [19].By combining outputs from multiple models,hybrid frameworks reduce prediction variance across different student populations [16].These approaches have shown to generalize better when applied to datasets that are collected from different academic contexts [19]. Recent literature has also highlighted the importance of explaining concepts in student performance prediction models [20].Explainable ML techniques help educators to understand the reasoning of the predictions made,and make predictions more reliable [20].Transparency is important in educational institutions as ethical considerations and accountability are important here [17]. Overall,contemporary research reflects a shift towards more intelligent,interpretable and scalable prediction frameworks for educational analytics [16].The combination of deep learning, hybrid modeling and explainable AI is a very a promising direction for future student performance prediction systems [18][20]. Several studies have emphasized the importance of se-lecting appropriate machine learning algorithms for student performance prediction [10].Supervised learning techniques like Decision Trees,SVM,Random Forest and Logistic Re-gression are widely adopted due to their strong classifica-tion capabilities [11].Ensemble approaches further enhance prediction stability by combining multiple weak learners to improve generalization performance [12].Feature selection and dimensionality reduction methods are also frequently applied to remove irrelevant attributes and improve computational efficiency [8].Improper model selection or overfitting may reduce the reliability of predictive outcomes and affect institu-tional decision-making [14]. Recent research has increasingly focused on integrating explainable AI techniques to improve transparency in academic prediction systems [1].Interpretabil-ity methods such as SHAP and LIME assist educators in understanding the contribution of different attributes toward performance outcomes [23],[24].The growing availability of large-scale educational datasets and learning management sys-tem logs has enabled more data-driven analysis of student behavior [9].However,challenges such as model bias,fairness concerns and scalability in real-world academic environments remain critical research gaps that require further investigation [28],[34]. Table I shows the different machine learning methods applied on educational data [1]-[15], including supervised models,ensemble techniques and analytical reviews. TABLE I SUMMARY OF RELATED WORK (REFERENCES [1]–[15]) Author & Year Study Focus ML Techniques Used Key Findings Dataset / Paper Link Ahmed et al. (2025) [1] Academic performance pre-diction with explainability Random XGBoost, Regression Forest, Logistic Interpretability improves de-cision making in institutions Paper Link https://www.nature.com/articles/ s41598-025-12353-4 Gul et al. (2025) [2] Comprehensive ML frame-work for prediction Random Forest, SVM, Logistic Regression Data-driven framework improves predictive performance Paper Link https://link.springer.com/article/10.1007/ s10791-025-09585-3 Rahman et al. (2025) [3] Systematic review of AI in education SVM, Random Forest, Naive Bayes Identifies research trends and gaps in performance predic-tion Paper Link https://scholar.google.com/scholar?q= Artificial+intelligence+in+education+A+ systematic+review+of+machine+learning+ for+predicting+student+performance+Rahman Buzducea et al. (2024) [4] ML for institutional demic decisions aca- Decision Tree, Random Forest ML supports institutional planning and academic interventions Paper Link https://www.mdpi.com/2076-3417/14/6/2412 E. Ahmed (2024) [5] Student performance predic-tion using ML SVM, Decision Tree, Random Forest, KNN Comparative evaluation of ML models for prediction Paper Link https://www.hindawi.com/journals/complexity/ 2024/5567124/ Munir et al. (2024) [6] AI integration in digital edu-cation Random Forest, Logistic Regression ML techniques enhance dig-ital learning analytics Paper Link https://www.sciencedirect.com/science/article/ pii/S2666920X24000146 Al Husaini & Shukor (2024) [7] Factors affecting academic performance Logistic Regression, De-cision Tree Behavioral and academic indicators influence performance Paper Link https://www.researchgate.net/publication/ 379428942 Yag?c? (2022) [8] Educational data mining pre-diction SVM, Random Forest, KNN, Naive Bayes High classification perfor-mance achieved using en-semble models Paper Link https://slejournal.springeropen.com/articles/ 10.1186/s40561-022-00192-z Rao & Kumar (2021) [9] Online course performance prediction Decision Tree, Random Forest LMS behavioural data im-proves prediction accuracy Paper Link https://uijrt.com/articles/v2/i11/ UIJRTV2I110007.pdf S¸ ekerog?lu et al. (2021) [10] Systematic review of prediction models ML SVM, Random Forest, Logistic Regression Highlights dataset imbalance and evaluation gaps Paper Link https://www.mdpi.com/2076-3417/11/16/7376 Albreiki et al. (2021) [11] Review of ML techniques for prediction SVM, Random Forest, Naive Bayes Summarizes ML algorithms used in education prediction Paper Link https://www.mdpi.com/2227-7102/11/9/552 Namoun & Alshanqiti (2020) [12] Learning analytics and pre-diction review Decision Tree, Logistic Regression Educational analytics improves early detection of performance risks Paper Link https://www.mdpi.com/2076-3417/10/1/237 Hashim (2020) [13] Supervised ML performance prediction model Decision Tree, SVM Supervised learning models classify academic outcomes effectively Paper Link https://iopscience.iop.org/article/10.1088/ 1757-899X/928/3/032019 Enughwure & Ogbise (2020) [14] ML applications in education review Decision Tree, Random Forest ML improves predictive ed-ucational analytics Paper Link https://www.irjet.net/archives/V7/i5/ IRJET-V7I51123.pdf Altabrawee (2019) [15] et al. Predicting student academic performance SVM, Naive Bayes ML models effectively pre-dict university student out-comes Paper Link https://journalofbabylon.com/index.php/JUB/ article/view/1515 III. METHODOLOGY The methodology used includes,collect ion of historical student dataset and applying machine learning algorithms on them for prediction. Eight Machine learning algorithms namely Linear Regression,Naive Bayes,KNN,Logistic Regression,Decision Tree,Random Forest,SVM and XGBoost were used for machine training.The performance of all models was evaluated using evaluation metrics. This workflow is used so that the most relevant attributes help to predict outcomes. The used methodology inculcates scalability and adaptability across various academic environ-ments.Additional features like learning styles and beahviour in real time can be included using preprocessing and model training.To minimize overfitting and build a strong model cross-validation strategies are used. This design help proper deployment in real-world academic settings.

References

[1] W. Ahmed et al., “Learning-based academic performance prediction with explainability for enhanced decision-making in educational institutions,” Discover Computing, 2025. https://link.springer.com [2] M. N. Gul et al., “Data driven decisions in education using a comprehen-sive machine learning framework for student performance prediction,” Discover Computing, 2025. https://link.springer.com [3] N. F. A. Rahman et al., “Artificial intelligence in education: A systematic review of machine learning for predicting student performance,” Journal of Advanced Research in Applied Sciences and Engineering Technology, 2025. https://akademiabaru.com [4] C.-A. Buzducea et al., “Machine learning in education: Predict-ing student performance and guiding institutional decisions,” 2024. https://www.researchgate.net [5] E. Ahmed, “Student performance prediction using machine learning algorithms,” 2024. https://www.researchgate.net [6] H. Munir, B. Vogel, and A. Jacobsson, “Artificial intelligence and ma-chine learning approaches in digital education: A systematic revision,” 2024. https://www.sciencedirect.com [7] Y. N. S. Al Husaini and N. S. A. Shukor, “Factors affecting students’ academic performance: A review,” 2024. https://www.sciencedirect.com [8] M. Yag?c?, “Educational data mining: Prediction of students’ aca-demic performance using machine learning algorithms,” 2022. https://doi.org/10.1186/s40561-022-00192-z [9] G. M. Rao and K. K. Kumar, “Students performance prediction in online courses using machine learning algorithms,” 2021. https://uijrt.com [10] B. S¸ ekeroglu et al., “Systematic literature review on machine learning and student performance prediction,” 2021. https://doi.org/10.3390/app112210907 [11] B. Albreiki et al., “A systematic literature review of student performance prediction using machine learning techniques,” 2021. https://doi.org/10.3390/educsci11090552 [12] A. Namoun and A. Alshanqiti, “Predicting student performance using data mining and learning analytics techniques,” 2020. https://doi.org/10.3390/app11010237 [13] A. S. Hashim, “Student performance prediction model based on super-vised machine learning algorithms,” 2020. https://doi.org/10.1088/1757-899X/928/3/032019 [14] A. A. Enughwure and M. E. Ogbise, “Application of machine learning methods to predict student performance,” 2020. https://www.irjet.net [15] H. Altabrawee et al., “Predicting students’ performance using machine learning techniques,” 2019. https://www.iasj.net [16] B. S¸ ekeroglu et al., “Student performance prediction and classification using machine learning algorithms,” 2019. https://dl.acm.org [17] A. Namoun, “Educational analytics and early warning systems for student performance prediction,” 2023. https://www.researchgate.net [18] S. M. Ajibade et al., “Educational data mining: Enhancement of student performance model using ensemble methods,” 2020. https://iopscience.iop.org [19] L. Breiman, “Random forests,” 2001. https://doi.org/10.1023/A:1010933404324 [20] C. Cortes and V. Vapnik, “Support-vector networks,” 1995. https://doi.org/10.1007/BF00994018 [21] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” 2016. https://doi.org/10.1145/2939672.2939785 [22] J. H. Friedman, “Greedy function approximation: A gradient boosting machine,” 2001. https://doi.org/10.1214/aos/1013203451 [23] S. Lundberg and S. Lee, “A unified approach to interpreting model predictions,” 2017. https://papers.nips.cc [24] M. Ribeiro et al., “Why should I trust you? Explaining the predictions of any classifier,” 2016. https://doi.org/10.1145/2939672.2939778 [25] I. Goodfellow et al., Deep Learning. https://www.deeplearningbook.org [26] T. Hastie et al., The Elements of Statistical Learning. https://hastie.su.domains/ElemStatLearn/ [27] F. Pedregosa et al., “Scikit-learn: Machine learning in Python.” https://jmlr.org/papers/v12/pedregosa11a.html [28] A. Barocas et al., Fairness and Machine Learning. https://fairmlbook.org [29] F. Trujillo, M. Pozo, and G. Suntaxi, “Artificial intelligence in education: A systematic literature review of machine learning approaches in student career prediction,” Journal of Technology and Science Education, vol. 15, no. 1, pp. 162–185, 2025. https://doi.org/10.3926/jotse.2603 [30] S. M. F. D. Syed Mustapha, “Predictive analysis of students’ learning performance using data mining techniques: A comparative study of feature selection methods,” Applied System Innovation, vol. 6, 2023. https://doi.org/10.3390/asi6030086 [31] P. Manjare, R. Shelke, and S. Kaware, “Comparative study of stu-dent academic outcomes and behavioral patterns through data-driven approaches,” ITM Web of Conferences, vol. 81, 2026. https://www.itm-conferences.org/ [32] A. Gole, S. Singh, P. Kanherkar, P. R. Abhishek, and P. P. Wankhede, “Comparative analysis of machine learning algorithms: Random Forest, Naive Bayes classifier and KNN – A survey,” International Journal for Research Publication & Seminar, vol. 13, no. 3, 2022. https://ijrps.org/ [33] Z. Sun et al., “Comparing machine learning models and statistical models for predicting heart failure events: A systematic review and meta-analysis,” Frontiers in Cardiovascular Medicine, vol. 9, 2022. https://doi.org/10.3389/fcvm.2022.863842 [34] T. P. Pagano et al., “Bias and unfairness in machine learning models: A systematic review on datasets, tools, fairness metrics, and mitigation methods,” Big Data and Cognitive Computing, vol. 7, no. 15, 2023. https://doi.org/10.3390/bdcc7010015 [35] D. Markovics and M. J. Mayer, “Comparison of machine learning methods for photovoltaic power forecasting based on numerical weather prediction,” Renewable and Sustainable Energy Reviews, vol. 161, 2022. https://doi.org/10.1016/j.rser.2022.112364 [36] T. Niu et al., “Decoding student cognitive abilities: A comparative study of explainable AI algorithms in educational data mining,” Scientific Reports, vol. 15, 2025. https://www.nature.com/ [37] Q. Long et al., “Enhancing happiness and life satisfaction in university students: Analysis with a machine learning approach,” BMC Psychology, vol. 14, 2026. https://bmcpsychology.biomedcentral.com/ [38] M. Jain and A. Srihari, “Comparison of machine learning algorithms in intrusion detection systems: A review using binary logistic regression,” International Journal of Computer Science and Mobile Computing, vol. 13, no. 10, pp. 45–53, 2024. https://ijcsmc.com/ [39] N. Al-Shanableh et al., “Forecasting students’ academic performance in educational data using machine learning techniques,” International Journal of Information and Communication Technology Education, vol. 22, no. 1, 2026. https://www.igi-global.com/ [40] J. K. Rogers, T. C. Mercado, and R. Cheng, “Predicting student performance using Moodle data and machine learning with feature im-portance,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 37, no. 1, pp. 223–231, 2025. https://ijeecs.iaescore.com/ [41] Y. Chen et al., “Machine learning-driven student performance prediction for enhancing tiered instruction,” arXiv preprint, 2025. https://arxiv.org/ [42] S. Wang et al., “Artificial intelligence in education: A systematic literature review,” Expert Systems with Applications, vol. 252, 2024. https://doi.org/10.1016/j.eswa.2024.124167 [43] M. A. Cardona, R. J. Rodr´?guez, and K. Ishmael, “Artificial intelligence and the future of teaching and learning: Insights and recommendations,” U.S. Department of Education, 2023. https://www.ed.gov/ [44] S. Somvanshi et al., “From tiny machine learning to tiny deep learning: A survey,” ACM Computing Surveys, 2026. https://dl.acm.org/ [45] M. Mart´?nez-Comesan˜a et al., “Impact of artificial intelligence on assessment methods in primary and secondary education: System-atic literature review,” Revista de Psicodida´ctica, vol. 28, 2023. https://doi.org/10.1016/j.psicoe.2022.11.001 [46] J. Wang and Y. Yu, “Machine learning approach to student perfor-mance prediction of online learning,” PLOS ONE, vol. 20, 2025. https://doi.org/10.1371/journal.pone.0300000 [47] G. Wei, G.-S. Han, and X. Lang, “Fire risk assessment using machine learning techniques: A case study of Jinan City, China,” Scientific Reports, 2026. https://www.nature.com/ [48] S. Albugami, A. Wali, and H. Almagrabi, “Predicting student dropout in Saudi universities using machine learning and explainable AI,” PeerJ Computer Science, 2026. https://peerj.com/ [49] A. M. Vieriu and G. Petrea, “The impact of artificial intelligence (AI) on students’ academic development,” Education Sciences, vol. 15, no. 3, 2025. https://doi.org/10.3390/educsci15030343 [50] A. I. AbuEid et al., “Artificial intelligence in education predicting college plans of high school students,” Journal of Intelligent System and Applied Data Science, vol. 2, no. 1, 2024. https://jisads.org

Copyright

Copyright © 2026 Om Deokar, Rajeev Patil, Vijaykumar P. Mantri. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.