scholarly journals ON-LINE NEO-PHASE AUTOENKODER FOR SYSTEMS WITH DEEP LEARNING ON THE BASE OF THE KOLMOGOROV’S NEURO-PHASE NETWORK

2017 ◽  
pp. 60-67
Author(s):  
Є.В. БОДЯНСЬКИЙ ◽  
О.А. ВИНОКУРОВА ◽  
Д.Д. ПЕЛЕШКО ◽  
Ю.М. РАШКЕВИЧ
Keyword(s):  
Neural Networks ◽  
Fuzzy System ◽  
Deep Neural Networks ◽  
Significant Loss ◽  
High Dimensional ◽  
Batch Mode ◽  
Neuro Fuzzy ◽  
Main Disadvantage ◽  
On Line ◽  
Inherent Part

One of the important problem, which is connected with big high dimensional data processing, is the task of their compression without significant loss of information that is contained in this data. The systems, which solve this problem and are called autoencoders, are the inherent part of deep neural networks. The main disadvantage of well-known autoencoders is low speed of learning process, which is implemented in the batch mode. In the paper the two-layered autoencoder is proposed. This system is the modification of Kolmogorov’s neuro-fuzzy system. Thus, in the paper the hybrid neo-fuzzy syste-  mencoder is proposed that has essentially advantages comparatively with conventional neurocompressors-encoders.

scholarly journals Efficient approximation of solutions of parametric linear transport equations by ReLU DNNs

2021 ◽  
Vol 47 (1) ◽  
Author(s):  
Fabian Laakmann ◽  
Philipp Petersen
Keyword(s):  
Neural Networks ◽  
Deep Neural Networks ◽  
Initial Conditions ◽  
Activation Function ◽  
Transport Equations ◽  
High Dimensional ◽  
Linear Transport ◽  
Approximation Rates ◽  
Curse Of Dimension ◽  
Efficient Approximation

AbstractWe demonstrate that deep neural networks with the ReLU activation function can efficiently approximate the solutions of various types of parametric linear transport equations. For non-smooth initial conditions, the solutions of these PDEs are high-dimensional and non-smooth. Therefore, approximation of these functions suffers from a curse of dimension. We demonstrate that through their inherent compositionality deep neural networks can resolve the characteristic flow underlying the transport equations and thereby allow approximation rates independent of the parameter dimension.

scholarly journals Experimental study of a novel neuro-fuzzy system for on-line handwritten UNIPEN digit recognition

1998 ◽  
Vol 19 (3-4) ◽  
pp. 357-364 ◽  
Author(s):  
E. Gómez Sánchez ◽  
J.A. Gago González ◽  
Y.A. Dimitriadis ◽  
J.M. Cano Izquierdo ◽  
J. López Coronado
Keyword(s):  
Experimental Study ◽  
Fuzzy System ◽  
Digit Recognition ◽  
Neuro Fuzzy ◽  
On Line

Intelligent Biometric System Using Soft Computing Tools

2011 ◽  
pp. 259-276
Author(s):  
Anupam Shukla ◽  
Ritu Tiwari ◽  
Chandra Prakash Rathore
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Soft Computing ◽  
Fuzzy System ◽  
Person Identification ◽  
Security Applications ◽  
Neuro Fuzzy ◽  
Biometric Systems ◽  
Physical Attributes ◽  
Novel Concept

Biometric Systems verify the identity of a claimant based on the person’s physical attributes, such as voice, face or fingerprints. Its application areas include security applications, forensic work, law enforcement applications etc. This work presents a novel concept of applying Soft Computing Tools, namely Artificial Neural Networks and Neuro-Fuzzy System, for person identification using speech and facial features. The work is divided in four cases, which are Person Identification using speech biometrics, facial biometrics, fusion of speech and facial biometrics and finally fusion of optimized speech and facial biometrics.

Neuro-Fuzzy System Modeling

2010 ◽  
pp. 147-175
Author(s):  
Chen-Sen Ouyang
Keyword(s):  
Neural Networks ◽  
Fuzzy System ◽  
Fuzzy Systems ◽  
System Modeling ◽  
Fuzzy Modeling ◽  
Structure Identification ◽  
Learning Capability ◽  
Computing Paradigm ◽  
Emerging Trends ◽  
Neuro Fuzzy

Neuro-fuzzy modeling is a computing paradigm of soft computing and very efficient for system modeling problems. It integrates two well-known modeling approaches of neural networks and fuzzy systems, and therefore possesses advantages of them, i.e., learning capability, robustness, human-like reasoning, and high understandability. Up to now, many approaches have been proposed for neuro-fuzzy modeling. However, it still exists many problems need to be solved. In this chapter, the authors firstly give an introduction to neuro-fuzzy system modeling. Secondly, some basic concepts of neural networks, fuzzy systems, and neuro-fuzzy systems are introduced. Also, they review and discuss some important literatures about neuro-fuzzy modeling. Thirdly, the issue for solving two most important problems of neuro-fuzzy modeling is considered, i.e., structure identification and parameter identification. Therefore, the authors present two approaches to solve these two problems, respectively. Fourthly, the future and emerging trends of neuro-fuzzy modeling is discussed. Besides, the possible research issues about neuro-fuzzy modeling are suggested. Finally, the authors give a conclusion.

scholarly journals Intrinsic motivation and episodic memories for robot exploration of high-dimensional sensory spaces

2020 ◽  
pp. 105971232092291
Author(s):  
Guido Schillaci ◽  
Antonio Pico Villalpando ◽  
Verena V Hafner ◽  
Peter Hanappe ◽  
David Colliaux ◽  
...  
Keyword(s):  
Neural Networks ◽  
Episodic Memory ◽  
Intrinsic Motivation ◽  
Computational Models ◽  
Deep Neural Networks ◽  
Image Sensor ◽  
Forward Kinematics ◽  
High Dimensional ◽  
Episodic Memories ◽  
Low Dimensional

This work presents an architecture that generates curiosity-driven goal-directed exploration behaviours for an image sensor of a microfarming robot. A combination of deep neural networks for offline unsupervised learning of low-dimensional features from images and of online learning of shallow neural networks representing the inverse and forward kinematics of the system have been used. The artificial curiosity system assigns interest values to a set of pre-defined goals and drives the exploration towards those that are expected to maximise the learning progress. We propose the integration of an episodic memory in intrinsic motivation systems to face catastrophic forgetting issues, typically experienced when performing online updates of artificial neural networks. Our results show that adopting an episodic memory system not only prevents the computational models from quickly forgetting knowledge that has been previously acquired but also provides new avenues for modulating the balance between plasticity and stability of the models.

Sign in / Sign up

Export Citation Format

Share Document