Dynamic Modification of Activation Function using the Backpropagation Algorithm in the Artificial Neural Networks
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2019, Vol 10, Issue 4
Abstract
The paper proposes the dynamic modification of the activation function in a learning technique, more exactly backpropagation algorithm. The modification consists in changing slope of sigmoid function for activation function according to increase or decrease the error in an epoch of learning. The study was done using the Waikato Environment for Knowledge Analysis (WEKA) platform to complete adding this feature in Multilayer Perceptron class. This study aims the dynamic modification of activation function has changed to relative gradient error, also neural networks with hidden layers have not used for it.
Authors and Affiliations
Marina Adriana Mercioni, Alexandru Tiron, Stefan Holban
Keywords
Related Articles
- EP ID EP550253
- DOI 10.14569/IJACSA.2019.0100406
- Views 90
- Downloads 0
Calculation of Pressure Loss Coefficients in Combining Flows of a Solar Collector using Artificial Neural Networks
The paper presents a novel technique for determination of loss coefficients due to pressure by use of artificial neural network (ANN) in tee junctions. Geometry and flow parameters are feed into ANN as the inputs for pur...
Hearing Aid Method by Equalizing Frequency Response of Phoneme Extracted from Human Voice
Hearing aid method by equalizing frequency response of phoneme which is extracted from human voice is proposed. One of the problems of the existing hearing aid is poor customization of the frequency response compensation...
A Feasibility Study on Porting the Community Land Model onto Accelerators Using Openacc
As environmental models (such as Accelerated Climate Model for Energy (ACME), Parallel Reactive Flow and Transport Model (PFLOTRAN), Arctic Terrestrial Simulator (ATS), etc.) became more and more complicated, we are faci...
Enhanced Random Early Detection using Responsive Congestion Indicators
Random Early Detection (RED) is an Active Queue Management (AQM) method proposed in the early 1990s to reduce the effects of network congestion on the router buffer. Although various AQM methods have extended RED to enha...
Texture Analysis on Image Motif of Endek Bali using K-Nearest Neighbor Classification Method
Endek fabric Bali is one form of craft woven fabric of Balinese society. Endek fabric has a variety of motifs or designs, a lot of people does not know that Endek have the type based on the design motif. In this research...