dc.contributorUniversidade Federal de São Carlos (UFSCar)
dc.contributorUniversidade Estadual Paulista (Unesp)
dc.date.accessioned2020-12-12T02:27:01Z
dc.date.accessioned2022-12-19T21:13:44Z
dc.date.available2020-12-12T02:27:01Z
dc.date.available2022-12-19T21:13:44Z
dc.date.created2020-12-12T02:27:01Z
dc.date.issued2019-07-01
dc.identifierProceedings of the International Joint Conference on Neural Networks, v. 2019-July.
dc.identifierhttp://hdl.handle.net/11449/201217
dc.identifier10.1109/IJCNN.2019.8851714
dc.identifier2-s2.0-85073157165
dc.identifier.urihttps://repositorioslatinoamericanos.uchile.cl/handle/2250/5381851
dc.description.abstractRestricted Boltzmann Machines achieved notorious popularity in the scientific community in the last decade due to outstanding results in a wide range of applications and also for providing the required mechanisms to build successful deep learning models, i.e., Deep Belief Networks and Deep Boltzmann Machines. However, their main bottleneck is related to the learning step, which is usually time-consuming. In this paper, we introduce a Sigmoid-like family of functions based on the Kaniadakis entropy formulation in the context of the RBM learning procedure. Experiments concerning binary image reconstruction are conducted in four public datasets to evaluate the robustness of the proposed approach. The results suggest that such a family of functions is suitable to increase the convergence rate when compared to standard functions employed by the research community.
dc.languageeng
dc.relationProceedings of the International Joint Conference on Neural Networks
dc.sourceScopus
dc.subjectKaniadakis Entropy
dc.subjectMachine Learning
dc.subjectRestricted Boltzmann Machines
dc.titleκ-Entropy Based Restricted Boltzmann Machines
dc.typeActas de congresos


Este ítem pertenece a la siguiente institución