Paper Info
Title
A New Data-Driven Deep Learning Model for Pattern Categorization using Fast Independent Component Analysis and Radial Basis Function Network. Taking Social Networks resources as a case.
Abstract
This paper investigates the categorization problem using Data Mining techniques. We present a new conceptual model, which is named FICARBFN, for classifying patterns by using Fast Fixed-Point Algorithm for Independent Component Analysis and Radial Basis Function Network. It uses an artificial neural network model to find a single consolidated categorization, which is composed of tree process, variables selection, categorization, and finally models selection. Our categorization model used a hybrid technique that combines the advantages of factorial analysis and Neural Network approaches. Comparative study and experimental results showed that our scheme optimized the bias-variance on the selected model and achieved an enhanced generalization for Social Networks patterns recognition.
Year
DOI
Venue
2017
10.1016/j.procs.2017.08.320
Procedia Computer Science
Keywords
Field
DocType
Data Mining,Pattern Recognition,Deep learning,Categorization,Variables Selection,Factor analysis
Data mining,Categorization,Radial basis function network,Social network,Data-driven,Conceptual model,Computer science,Independent component analysis,Artificial intelligence,Deep learning,Artificial neural network,Machine learning
Conference
Volume
ISSN
Citations 
113
1877-0509
0
PageRank 
References 
Authors
0.34
5
2
Name
Order
Citations
PageRank
Choukri Djellali154.48
Mehdi Adda25815.42