Controlling the Behavior of a Neural Network Weights Using Variables Correlation and Posterior Probabilities Estimation
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2014, Vol 16, Issue 3
Abstract
ABSTRACT: In this article, a number of posterior probabilistic based equations were introduced to detect the effect of controlling the correlation between variables on the behavior of feed forward neural network weights. In this paper it was proofed that, under certain assumptions, in a feed forward neural network with backprobagation learning algorithm, the correlation between the input variables on one side and the target variable on the other, is directly proportional to the values of the connection weights from the input layer to the output layer through the hidden layer.
Authors and Affiliations
Hazem Migdady, , Yousef Talafha, , Hussam Alrabaiah
Keywords
Related Articles
- EP ID EP147416
- DOI 10.9790/0661-16323641
- Views 122
- Downloads 0
Protection of Direct and Indirect Discrimination using Prevention Methods
Along with privacy, discrimination is a very important issue when considering the legal and ethical aspects of data mining. It is more than observable that the majority people do not want to be discriminated &nbs...
Assessing Buffering with Scheduling Schemes in a QoS Internet Router
A key requirement for service differentiation as required in Internet of the future and QoS to work effectively is the extension of traffic management routines of the current TCP/IP protocol. Two among of such &n...
Neural Networks Optimization, Using the Genetic Algorithm
Abstract: There are many ways in neural networks optimization by genetic algorithm. One of the optimization methods is the neural networks repetition. The main idea of this method is that the model is developed at first...
Automated system for deployment of websites and windows services to the production servers
Abstract: This paper is discussed on automated system for deployment of websites and windows services to the production servers. The aim of this paper is to develop and implement an automatic system for deployment...
Occlusion detection in video sequences
Abstract: An occlusion is the region between two overlapping objects with disparate motion. Detecting these occluded objects is crucial for many of the video processing. The Occlusion detection is decomposed into two ind...