Predicting Students' Progression in Higher Education by Using the Random Forest Algorithm

Hardman, Julie; Paucar‐Caceres, Alberto; Fielding, Alan H.

doi:10.1002/sres.2130

Cited by 33 publications

(11 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We used the machine learning technique 'Random Forest' (Breiman 2001) to identify the importance for specific survey instrument questions (Q1 through Q21) in predicting student 'Overall Satisfaction' (Q22). Machine learning methods are increasingly being used to analyse complex, large datasets in diverse fields of study, including psychology (Strobl et al 2009), ecology (Prasad et al 2006) and higher education research (Langan et al 2016;Hardman et al 2013). Random Forest analysis ultimately makes predictions based on variable associations, ranking the predictive importance of each variable (Grömping 2009;Genuer et al 2010).…”

Section: Machine Learning Analysismentioning

confidence: 99%

National student survey metrics: where is the room for improvement?

Langan

Harris

2019

High Educ

View full text Add to dashboard Cite

The purpose of this study is to use machine learning and exploratory data analysis to interrogate patterns of metrics from a national-level student survey. Analysis of over 1.8 million returns detected long-term stability of the predictors of student satisfaction, with survey items relating to course management and teaching being consistently most influential. All metric outputs increased over the survey period; however, the rates of increase of several dimensions including Overall Satisfaction decreased markedly in the most recent years to a point of levelling off. There was also a growing similarity in an institution of outcomes at a national level. This study contributes new insights into the influential survey instrument, through rigorous determination of the most influential survey items, descriptions of the changes in variability between institutions, and exploration of the importance of patterns of outliers at the extremes of the metric outputs. We also identify a rapidly growing spike in total satisfaction at a broad course level and highlight how this is inconsistent with a customer satisfaction model. We conclude by considering the challenges of the use of national-level student surveys for the management of student satisfaction in higher education.

show abstract

Section: Machine Learning Analysismentioning

confidence: 99%

National student survey metrics: where is the room for improvement?

Langan

Harris

2019

High Educ

View full text Add to dashboard Cite

show abstract

“…Hardmann et al . () highlighted the usefulness of random forest algorithm, a technique initially developed in biology, in evaluating an educational and learning environment. As researchers indicate that the education system represents an extremely complicated system that involves many aspects; thus, by using systems science methods (Lin et al .…”

Section: Discussionmentioning

confidence: 99%

“…Ho (2013) indicated recently that system concepts including cognitive system concepts, soft systems methodology, and soft systems thinking can be applied to e-learning systems. Hardmann et al (2013) highlighted the usefulness of random forest algorithm, a technique initially developed in biology, in evaluating an educational and learning environment. As researchers indicate that the education system represents an extremely complicated system that involves many aspects; thus, by using systems science methods (Lin et al 2013), we can analyse the complexity involved (Jackson, 1995).…”

Section: Discussionmentioning

confidence: 99%

Impact of Effective Communication, Achievement Sharing and Positive Classroom Environments on Learning Performance

Shan

Ji-hong

et al. 2014

Syst. Res.

View full text Add to dashboard Cite

Finding a way to improve students' learning performance is a crucial concern in education reform. Many researchers have investigated possible predictors of student learning performance, including personality traits, learning styles, self‐efficacy, and so on. Nevertheless, there is limited research that examines the effectiveness of the interaction between students and teachers, achievement sharing, and the classroom environment. This study investigates the impact of effective communication between students and teachers, achievement sharing, and a positive classroom environment on learning performance through an empirical analysis. We found that when teachers create a sense of community, respond to students and foster positive relationships, students are more engaged and enthusiastic about learning and tend to perform better academically. Copyright © 2014 John Wiley & Sons, Ltd.

show abstract

“…These strategies were applied to select the Base-learners in the first layer. Random forest (RF) [23][24][25], SVM [26][27][28], and AdaBoost [29,30] classifiers were chosen in this study to be the Base-learners due to their better classification performances. They are commonly adopted as predictive models for student achievement prediction at present and have good performance.…”

Section: The Proposed Ensemble Modelmentioning

confidence: 99%

An Ensemble Prediction Model for Potential Student Recommendation Using Machine Learning

Yan

Liu

2020

Symmetry

View full text Add to dashboard Cite

Student performance prediction has become a hot research topic. Most of the existing prediction models are built by a machine learning method. They are interested in prediction accuracy but pay less attention to interpretability. We propose a stacking ensemble model to predict and analyze student performance in academic competition. In this model, student performance is classified into two symmetrical categorical classes. To improve accuracy, three machine learning algorithms, including support vector machine (SVM), random forest, and AdaBoost are established in the first level and then integrated by logistic regression via stacking. A feature importance analysis was applied to identify important variables. The experimental data were collected from four academic years in Hankou University. According to comparative studies on five evaluation metrics (precision, recall, F1, error, and area under the receiver operating characteristic curve ( AUC ) in this analysis, the proposed model generally performs better than compared models. The important variables identified from the analysis are interpretable, they can be used as guidance to select potential students.

show abstract

Predicting Students' Progression in Higher Education by Using the Random Forest Algorithm

Cited by 33 publications

References 7 publications

National student survey metrics: where is the room for improvement?

National student survey metrics: where is the room for improvement?

Impact of Effective Communication, Achievement Sharing and Positive Classroom Environments on Learning Performance

An Ensemble Prediction Model for Potential Student Recommendation Using Machine Learning

Contact Info

Product

Resources

About