Publication: Development of a new hybrid regression model: an application of fuzzy regression and multilayer feedforward neural network
| dc.contributor.author | Tabnjh, Abedelmlek Kalefh Sleeman | |
| dc.date.accessioned | 2024-01-30T01:41:47Z | |
| dc.date.available | 2024-01-30T01:41:47Z | |
| dc.date.issued | 2023-08 | |
| dc.description.abstract | Herbal extracts have been utilized in oral health to treat various ailments, including inflammation, as antimicrobial plaque agents, antiseptics, antioxidants, histamine release prevention, and as antibacterial, antifungal, antiviral, and antimicrobial analgesics. Herbal medication also functions in healing, plaque reduction in the oral cavity, and immune enhancement. There is minimal research that has used regression to link knowledge with practice and other sociodemographic variables in HMOH. To develop a hybrid model by considering bootstrap, neural network, and fuzzy regression for HMOH KP, to measure the efficacy and efficiency of the developed hybrid model for HMOH KP, and to validate the newly developed hybrid model. This study aims to develop the best strategy for handling data analysis, especially in HMOH KP, which combines fuzzy regression and Multi-layer Feedforward Neural Network (MLFFNN). R-programming software is used to write the developed syntax. All the essential steps are summarized in the R syntax. The new hybrid regression model incorporating bootstrapping, MLFFNN, and fuzzy regression increases the precision of the estimated parameters and compensates for the ambiguous relationship between the dependent and independent variables. The MLFFNN method has successfully measured the effectiveness, efficiency, and accuracy of the new hybrid model. The R2 value and the predicted value obtained are used to validate the derived model. Conclusion: This thesis presents a new methodology for creating precise and validated regression models through the utilization of the HMOH KP dataset. Moreover, this approach can be extended to any other dataset that aligns with the provided assumptions. | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/18216 | |
| dc.language.iso | en | |
| dc.title | Development of a new hybrid regression model: an application of fuzzy regression and multilayer feedforward neural network | |
| dc.type | Resource Types::text::thesis::doctoral thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |