The research sought to investigate the effects of altering values of three selected hyperparameters (Epoch, Learning rate and Drop out), on the accuracy level of a deep belief network algorithm based web phishing detection system. The accuracy level of the system changed as these optimiser values were altered. The experimental results showed that there was a positive linear relationship between Epoch and learning rate values with regards to the accuracy of the system. The study identified a negative linear relationship between dropout values and accuracy of the web phishing detection system. The study hence recommended that keeping Epoch values high around or a little above 25 cycles, maintaining a lowdrop out value (0.1), while increasing the value of learning rate (>=0.7) would ensure arrival at a high accuracy level of the deep belief network algorithm based web phishing detection system (0.987).
Chinyere, C. O. C., & Angela, A. I. (2020). Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System. Transdisciplinary Agora for Future Discussions Journal, 2(2), 12–18. https://journal.taffds.org/articles/effect-of-selected-hyper-parameters-on-accuracy-level-of-deep-belief-network-algorithm-based-web-phishing-detection-system-2-12
Chinyere, Chigozie-Okwum Chioma, and Ajah Ifeyinwa Angela. “Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System.” Transdisciplinary Agora for Future Discussions Journal, vol. 2, no. 2, 2020, pp. 12–18. https://journal.taffds.org/articles/effect-of-selected-hyper-parameters-on-accuracy-level-of-deep-belief-network-algorithm-based-web-phishing-detection-system-2-12.
Chinyere, Chigozie-Okwum Chioma and Ajah Ifeyinwa Angela. 2020. “Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System.” Transdisciplinary Agora for Future Discussions Journal 2 (2): 12–18. https://journal.taffds.org/articles/effect-of-selected-hyper-parameters-on-accuracy-level-of-deep-belief-network-algorithm-based-web-phishing-detection-system-2-12.
Chinyere, C.O.C. and Angela, A.I. (2020) ‘Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System’, Transdisciplinary Agora for Future Discussions Journal, 2(2), pp. 12–18. Available at: https://journal.taffds.org/articles/effect-of-selected-hyper-parameters-on-accuracy-level-of-deep-belief-network-algorithm-based-web-phishing-detection-system-2-12 (Accessed: 15 October 2026).
C. O. C. Chinyere and A. I. Angela, “Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System,” Transdisciplinary Agora for Future Discussions Journal, vol. 2, no. 2, pp. 12–18, Aug. 2020.
Chinyere COC, Angela AI. Effect of Selected Hyper Parameters on Accuracy Level of Deep Belief Network Algorithm Based Web Phishing Detection System. TAFFD's J. 2020;2(2):12-18. Available from: https://journal.taffds.org/articles/effect-of-selected-hyper-parameters-on-accuracy-level-of-deep-belief-network-algorithm-based-web-phishing-detection-system-2-12
© 2020 The Author(s). Published by Transdisciplinary Agora for Future Discussions Inc. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.