|
Efficiency Comparison of Two TCAS Minimum Mark PredictionModels: Top X% Model and Mean+N(s.d.) Model forAdmission Round of TCAS in Year 2023 |
|---|---|
| รหัสดีโอไอ | |
| Creator | Padungsak Kasatecharoen |
| Title | Efficiency Comparison of Two TCAS Minimum Mark PredictionModels: Top X% Model and Mean+N(s.d.) Model forAdmission Round of TCAS in Year 2023 |
| Contributor | Ruethaichanok Kasatecharoen |
| Publisher | Thammasat University |
| Publication Year | 2567 |
| Journal Title | Thai Journal of Science and Technology |
| Journal Vol. | 12 |
| Journal No. | 2 |
| Page no. | 52-66 |
| Keyword | TCAS Minimum MarkPrediction, Top X% Model, Mean+ N(s.d.) Model, TCAS, Admission Round |
| URL Website | https://www.tci-thaijo.org/ |
| Website title | THAIJO |
| ISSN | 2286-7333 |
| Abstract | Thai university Central Admission System (TCAS) is developed for all high school students in Thailand. There are four rounds of TCAS called, portfolio, quota, admission and direct admission.Theadmission round is the round that use the mark of each test subject to compute the total mark, and then apply to TCAS. The minimum mark that can qualify the admission is the main point.However, the last year minimum mark cannot be used directly because of changing of subject combination, changing of fraction of the subject combination and the difficulty of the tests in each year. Otherwise, in year 2022, there are major changes in test subjects of TCAS. These lead to the prediction is more difficulty.The use of last year minimum marks directly is not efficient method because of its low accuracy. The researchers have developed two models calledโTop X% modelโand โMean + N(s.d.) modelโto predict the minimum mark. This paper represents the concept and methodology of each model, each model efficiency and compare their efficiency. The results show clearly that โTop X% modelโ is better than โMean + N(s.d.) modelโ in term of model accuracy and error prediction average. The results also show that the prediction accuracy of โTop X% modelโ and โMean + N(s.d.) modelโ for all four universities in year 2023 are 78.6744% and 75.5043%, respectively, with acceptance of the prediction total mark error not exceed 5%. |
