Comparing classification models in the final exam performance prediction

Gamulin, Jasna; Gamulin, Ozren; Kermek, Dragutin

doi:10.1109/mipro.2014.6859650

Cited by 7 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Those results are confirming the assumption that reducing the number of coefficients in periodogram to the smaller number of most commonly (frequently) used coefficients will not degrade information about observed student and on the other side, according to Figure (c), will enable time‐efficient model calculation. Model accuracies for all used algorithms listed in Tables and are comparable with accuracies found in literature sources (Kotsiantis et al ., ; Delgado et al ., ; Thai‐Nghe et al ., ; Romero et al ., ; Lykourentzou et al ., ; Kotsiantis et al ., ; Jovanović et al ., ; Gamulin et al ., ; Lara et al ., ) confirming that idea of using Fourier coefficients for student success prediction is plausible. Small differences between courses can be ascribed to the different numbers of students, student type (graduate, undergraduate, technology oriented or not), quantity and quality of resources posted at course page and so on.…”

Section: Resultssupporting

confidence: 69%

“…In all models, the split validation with stratified sampling was performed, meaning that one set of data was used for model building (70%) and the other for validation of the model (30%). During the model calculation, attributes were optimized and selected with genetic algorithm because it was shown (Oreški et al ., ; Gamulin et al ., ) that the model accuracy was much higher when optimization was applied.…”

Section: Methodology and Datasetmentioning

confidence: 96%

See 1 more Smart Citation

Using Fourier coefficients in time series analysis for student performance prediction in blended learning environments

Gamulin

Kermek

2015

Expert Systems

Self Cite

View full text Add to dashboard Cite

In this work, it is shown that student access time series generated from Moodle log files contain information sufficient for successful prediction of student final results in blended learning courses. It is also shown that if time series is transformed into frequency domain, using discrete Fourier transforms (DFT), the information contained in it will be preserved. Hence, resulting periodogram and its DFT coefficients can be used for generating student performance models with the algorithms commonly used for that purposes. The amount of data extracted from log files, especially for lengthy courses, can be huge. Nevertheless, by using DFT, drastic compression of data is possible. It is experimentally shown, by means of several commonly used modelling algorithms, that if in average all but 5–10% of most intensive and most frequently used DFT coefficients are removed from datasets, the modelling with the remained data will result with the increase of the model accuracy. Resulting accuracy of the calculated models is in accordance with results for student performance models calculated for different dataset types reported in literature. The advantage of this approach is its applicability because the data are automatically collected in Moodle logs.

show abstract

Section: Resultssupporting

confidence: 69%

Section: Methodology and Datasetmentioning

confidence: 96%

Using Fourier coefficients in time series analysis for student performance prediction in blended learning environments

Gamulin

Kermek

2015

Expert Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…The Area Under the ROC Curve (AUC) is a universal statistical indicator for describing the accuracy of a model regarding predicting a phenomenon [32]. It has been widely used in education research for comparing classification algorithms and models [33,34] instead of other well-known evaluation measures such as Accuracy, F-measure, Sensitivity, Precision, etc. AUC can be defined as the probability that a classifier will rank a randomly chosen positive instance higher than a randomly chosen negative one (assuming 'positive' ranks higher than 'negative').…”

Section: Experimentationmentioning

confidence: 99%

Towards Portability of Models for Predicting Students’ Final Performance in University Courses Starting from Moodle Logs

López-Zambrano

Lara²,

Romero

2020

Applied Sciences

View full text Add to dashboard Cite

Predicting students’ academic performance is one of the older challenges faced by the educational scientific community. However, most of the research carried out in this area has focused on obtaining the best accuracy models for their specific single courses and only a few works have tried to discover under which circumstances a prediction model built on a source course can be used in other different but similar courses. Our motivation in this work is to study the portability of models obtained directly from Moodle logs of 24 university courses. The proposed method intends to check if grouping similar courses by the degree or the similar level of usage of activities provided by the Moodle logs, and if the use of numerical or categorical attributes affect in the portability of the prediction models. We have carried out two experiments by executing the well-known classification algorithm over all the datasets of the courses in order to obtain decision tree models and to test their portability to the other courses by comparing the obtained accuracy and loss of accuracy evaluation measures. The results obtained show that it is only feasible to directly transfer predictive models or apply them to different courses with an acceptable accuracy and without losing portability under some circumstances.

show abstract

“…One of the research which has adapted machine learning algorithms has shown that early prediction of students' grades might help to improve www.aetic.theiaer.org students' success rate, result and course selection. Gamulin et al [2] predicts students' performance for the final examination based on the machine learning and it has successfully improved the learning process. In the research, the predictions have given the chance for students and teachers to improve based on their predicted results.…”

Section: Introductionmentioning

confidence: 99%

Prediction of MUET Results Based on K-Nearest Neighbour Algorithm

Sabri,

Hamrizan

2023

AETiC

View full text Add to dashboard Cite

The machine learning based prediction has been applied in various fields to solve different kind of problems. In education, the research on the predictions of examination results is gaining more attentions among the researchers. The adaptation of machine learning for the prediction of students’ achievement enables the educational institutions to identify the high failure rate, learning problems, and reasons for low student performance. This research is proposing the prediction of the Malaysian University English Test (MUET) results based on the K-Nearest Neighbour Algorithm (KNN). KNN is a powerful algorithm that has been applied in various prediction problems. The prediction of the MUET results would help the students and lecturers to be more well prepared and could improve the required English language skills accordingly before the actual examination. The MUET result prediction is based on the student’s English courses grades and there are 516 data of students’ results that have been collected from Universiti Teknologi MARA (UiTM) Dungun campus. The performance measurement that has been used are the mean accuracy, percentage error and mean squared error (MSE). In this research, the KNN prediction model has generated an acceptable performance with 65.29% accuracy. For future work, KNN could be modified or hybridized to further improve its performance. Furthermore, other algorithms could also be explored into this problem to further validate the best predictive model for the prediction of the MUET results.

show abstract

Comparing classification models in the final exam performance prediction

Cited by 7 publications

References 12 publications

Using Fourier coefficients in time series analysis for student performance prediction in blended learning environments

Using Fourier coefficients in time series analysis for student performance prediction in blended learning environments

Towards Portability of Models for Predicting Students’ Final Performance in University Courses Starting from Moodle Logs

Prediction of MUET Results Based on K-Nearest Neighbour Algorithm

Contact Info

Product

Resources

About