Synergizing intelligence and knowledge discovery: Hybrid black hole algorithm for optimizing discrete Hopfield neural network with negative based systematic satisfiability DOI Creative Commons
Nur ‘Afifah Rusdi, Nur Ezlin Zamri, Mohd Shareduwan Mohd Kasihmuddin

и другие.

AIMS Mathematics, Год журнала: 2024, Номер 9(11), С. 29820 - 29882

Опубликована: Янв. 1, 2024

<p>The current systematic logical rules in the Discrete Hopfield Neural Network encounter significant challenges, including repetitive final neuron states that lead to issue of overfitting. Furthermore, neglect impact on appearance negative literals within structure, and most recent efforts have primarily focused improving learning capabilities network, which could potentially limit its overall efficiency. To tackle limitation, we introduced a Negative Based Higher Order Systematic Logic imposing restriction clauses. Additionally, Hybrid Black Hole Algorithm was proposed retrieval phase optimize states. This ensured optimized achieved maximum diversity reach global minima solutions with lowest similarity index, thereby enhancing performance network. The results illustrated model can achieve up 10,000 diversified an average index 0.09. findings indicated are optimal configurations. findings, development new SAT implementation algorithm DHNN multi-objective functions result updated high diversity, attainment solutions, produces low index. Consequently, this be extended for logic mining applications classification tasks. will enhance high-quality induced logic, is effective knowledge extraction.</p>

Язык: Английский

Synergizing intelligence and knowledge discovery: Hybrid black hole algorithm for optimizing discrete Hopfield neural network with negative based systematic satisfiability DOI Creative Commons
Nur ‘Afifah Rusdi, Nur Ezlin Zamri, Mohd Shareduwan Mohd Kasihmuddin

и другие.

AIMS Mathematics, Год журнала: 2024, Номер 9(11), С. 29820 - 29882

Опубликована: Янв. 1, 2024

<p>The current systematic logical rules in the Discrete Hopfield Neural Network encounter significant challenges, including repetitive final neuron states that lead to issue of overfitting. Furthermore, neglect impact on appearance negative literals within structure, and most recent efforts have primarily focused improving learning capabilities network, which could potentially limit its overall efficiency. To tackle limitation, we introduced a Negative Based Higher Order Systematic Logic imposing restriction clauses. Additionally, Hybrid Black Hole Algorithm was proposed retrieval phase optimize states. This ensured optimized achieved maximum diversity reach global minima solutions with lowest similarity index, thereby enhancing performance network. The results illustrated model can achieve up 10,000 diversified an average index 0.09. findings indicated are optimal configurations. findings, development new SAT implementation algorithm DHNN multi-objective functions result updated high diversity, attainment solutions, produces low index. Consequently, this be extended for logic mining applications classification tasks. will enhance high-quality induced logic, is effective knowledge extraction.</p>

Язык: Английский

Процитировано

0