Matching preprocessing methods for improving the prediction of student's graduation

Prachuabsupakij, Wanthanee; Doungpaisan, Pafan

doi:10.1109/compcomm.2016.7924659

Cited by 2 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposed model for generating detection rulesets is evaluated using a confusion matrix to measure the model's performance regarding Precision, Recall, F-Measure, and G-Mean [30]. F-Measure is used to measure minority rates in imbalanced classes, and the G-mean index is used to measure overall classification performance.…”

Section: Generating Rules and Model Evaluationmentioning

confidence: 99%

A Rule-Generation Model for Class Imbalances to Detect Student Entrepreneurship Based on the Theory of Planned Behavior

Rijati

Purwitasar

Sumpeno

et al. 2022

Cybernetics and Information Technologies

View full text Add to dashboard Cite

The ability to identify the entrepreneurial potential of students enables higher education institutions to contribute to the economic and social development of a country. Current research trends regarding the detection of student entrepreneurial potential have the greatest challenge in the unequal ratio of datasets. This study proposes a rule-generation model in an imbalanced situation to classify student entrepreneurship based on the Theory of Planned Behavior (TPB). The result is a ruleset that is used for the early detection of student entrepreneurial potential. The proposed method consists of three main stages, namely preprocessing data to classify data based on TPB variables, generating a dataset by clustering and selecting attributes by sampling to balance the data, and finally generating a ruleset. Furthermore, the results of the detecting ruleset have been evaluated with actual data from the student tracer study as ground truth. The evaluation results show high accuracy so that the ruleset can be applied to the higher education environment in the future.

show abstract

Section: Generating Rules and Model Evaluationmentioning

confidence: 99%