Roslan, Nurshazneem (2025) Conditional Random 2 Satisfiability In Discrete Hopfield Neural Networks For Enhanced Logic Mining. PhD thesis, Perpustakaan Hamzah Sendut.
![[img]](http://eprints.usm.my/style/images/fileicons/application_pdf.png) |
PDF
Download (519kB) |
Abstract
A deeper understanding of additional mechanisms in discrete hopfield neural network is essential for developing intelligent model for broader applications. This thesis introduces conditional random 2 satisfiability, a new logical rule guaranteeing the inclusion of at least one negative logic operator for each second-order clause which implements the non-monotonic smish activation function to enhance the updating process within the network. Furthermore, multi-objective function optimization using a hybrid binary whale optimization algorithm is employed during the retrieval phase to generate diversified neuron states with maximum global solutions and minimized similarity indices. Finally, a new approach for identifying the best logic based on various performance metrics in the logic mining model called conditional random 2 satisfiability reverse analysis is proposed. This approach is significant when dealing with imbalanced datasets leading to an enhanced search space for finding optimal induced logic.
| Item Type: | Thesis (PhD) |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA1 Mathematics (General) |
| Divisions: | Pusat Pengajian Sains Matematik (School of Mathematical Sciences) > Thesis |
| Depositing User: | Mr Hasmizar Mansor |
| Date Deposited: | 24 Jun 2026 00:42 |
| Last Modified: | 24 Jun 2026 00:42 |
| URI: | http://eprints.usm.my/id/eprint/64432 |
Actions (login required)
 |
View Item |
