A Procedure to Select the Vigilance Threshold for the ART2 for Supervised and Unsupervised Training

Este artículo describe el diseño de un software educativo (Body Ar) también aplicación móvil basada en realidad aumentada y enfocada a la educación básica, específicamente, sobre la anatomía humana que se estudia en tercer grado de primaria en la asignatura de Ciencias Naturales. Las principales características propuestas en el diseño del software es que los niños aprendan datos interesantes de una forma divertida, y entretenida, sobre cómo funciona el cuerpo humano por medio de la visualización de imágenes en tercera dimensión. Otra característica de Body Ar es brindarle al usuario la posibilidad de evaluar los conocimientos obtenidos conforme vaya avanzando con los temas proporcionados por medio de distintas actividades que le permitan al mismo tiempo reafirmar lo que va aprendiendo.

show abstract

“…De acuerdo con Rayón y Sossa [2] existen distintos tipos de software educativo que se clasifican de la siguiente manera:…”

Section: Introductionunclassified

Body Ar: Sistema educativo de anatomía basado en realidad aumentada

Sanchez¹,

Torres²,

Valdez³

et al. 2019

Progmat

View full text Add to dashboard Cite

show abstract

“…For the first group, scientists tried to optimize learning algorithms of original ART-2 network to obtain better results while keeping the network structure unchanged. For example, Villela and Azuela discussed how to apply ART-2 neural network to the pattern classification problem using a set of sub-feature vectors selected artificially, and the authors also proposed a method to select the vigilance threshold in ART-2 based on a predefined training dataset [21]. Qian used a single ART-2 neural network to realize multi-layer dynamic clustering by simultaneously considering the information from both the winning neuron and other neurons in competition as well as Hebb rule [22].…”

Section: Introductionmentioning

confidence: 99%

Robust ART-2 neural network learning framework

Yin

Shen

2011

Proceedings of 2011 International Conference on Modelling, Identification and Control

View full text Add to dashboard Cite

The ART-2 network is a typical adaptive resonance theory based neural network approach for clustering purpose and has been successfully used in many fields. However, one of the fatal shortcomings of traditional ART-2 is that its final results heavily depend on a pre-defined fixed vigilance threshold parameter, which makes it infeasible to be applied in different complicated applications. Another disadvantage of traditional ART-2 method is that the number of categories in the network will increase all the time with the continuous input. Considering these points, an improved algorithm of ART-2 has been presented in this paper called the Robust ART-2. We first systematically analyze the dynamic changes of the optimal vigilance threshold with the succession inputs and propose a new adaptive method to make the network itself can automatically choose the optimal threshold in various situations. Then we introduce a constraint parameter to confine the scale of ART-2 network by limiting the maximal number of categories of network. Simulation experiments including artificial and benchmark data sets demonstrate the effectiveness of our algorithm.

show abstract

Improving Pattern Recognition Using Several Feature Vectors

Villela

Azuela²

2002

MICAI 2002: Advances in Artificial Intelligence

View full text Add to dashboard Cite

A Procedure to Select the Vigilance Threshold for the ART2 for Supervised and Unsupervised Training

Cited by 4 publications

References 1 publication

Body Ar: Sistema educativo de anatomía basado en realidad aumentada

Body Ar: Sistema educativo de anatomía basado en realidad aumentada

Robust ART-2 neural network learning framework

Improving Pattern Recognition Using Several Feature Vectors

Contact Info

Product

Resources

About