Publication: Alternative method to develop new strategy in ordinal regression: a case study in dental
| dc.contributor.author | Lazin, Muhamamd Amirul Mat | |
| dc.date.accessioned | 2025-09-07T07:57:04Z | |
| dc.date.available | 2025-09-07T07:57:04Z | |
| dc.date.issued | 2025-03 | |
| dc.description.abstract | Clinical data usually contain numerous features with a small sample size, resulting in higher dimensionality and poor accuracy. This reduces the performance of classifier systems in high-dimensional data sets because irrelevant features contribute to poor classification accuracy and add extra difficulties in finding potentially useful knowledge. The main objective is to develop an alternative model for ordinal regression through statistical methodology building. The methodology includes a computational study design and statistical techniques customised for dental science modelling. A combination of ordinal regression and bootstrap techniques in the developing an alternative model is the main key to the research focal point. Two case studies, tooth wear severity and tooth sensitivity, were used to test this technique, demonstrating its relevance to real-world dental data. All the fundamental programming was performed using R software. The results show that the alternative approach, especially with more bootstrap replications, offers improved model fitting and precision compared to traditional ordinal regression. This suggests its usefulness in improving the accuracy of health science research, especially in situations with small sample sizes. This study strengthens statistical methods in dental sciences by introducing a more robust alternative to ordinal regression, enabling researchers to obtain more accurate and reliable results even with limited datasets. | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/22503 | |
| dc.language.iso | en | |
| dc.title | Alternative method to develop new strategy in ordinal regression: a case study in dental | |
| dc.type | Resource Types::text::thesis::master thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |