Neuro-Fuzzy Hybrid System in Soft Computing
INTRODUCTION:
Neural Network: - An neural network in general, is a highly interconnected network of a large number of processing elements called Neurons in an architecture inspired by the human brain.
u The objective of a neural network is to transform input into meaningful output.
u Neural network learns by examples.
Explanation of Neuro-Fuzzy system: https://www.youtube.com/watch?v=mV5vNaXypwc&feature=youtu.be
Explanation of Neuro-Fuzzy system: https://www.youtube.com/watch?v=mV5vNaXypwc&feature=youtu.be
Neuro-Fuzzy Hybrid System (Part-2)- Soft Computing ~xRay Pixy
What is Neuro-Fuzzy Hybrid System - Soft Computing
Explanation of Neuro Fuzzy Hybrid System
Fuzzy Logic or Fuzzy Set Theory: - Fuzzy means not clear, distinct, precise or blurred (with unclear outline).
u It is a flexible machine learning technique.
u Fuzzy logic deals with uncertainty or vagueness existing in a system and formulating fuzzy rules to find a solution to problems.
u Fuzzy logic use values between 0 and 1.
u Fuzzy set also consist of Fuzzy rule base to perform approximate reasoning somewhat similar to the human brain.
u Example of Fuzzy logic: - For automatic breaking system the traditional values are taken as either 0 or 1.
Central driving force for the creation of hybrid soft computing systems:
Every soft computing technique has particular computational parameters which make them suited for a particular problem and not for others.
o ability to learn & decision making
u Neural Networks are good at recognizing patterns but they are not good at explaining how they reach their decisions.
u Fuzzy logic is good at explaining the decisions but cannot automatically acquire the rules used for making the decision.
u These limitations act as a central driving force for the creation of hybrid soft computing systems where two or more techniques are combined in a suitable manner that overcomes the of individual techniques.
u The aim is to build highly automated, intelligent machines for the future generations using all of these techniques.
Hybrid System: - A Hybrid Intelligent System is one that combines at least two intelligent technologies.
u For example, combining a neural network with a fuzzy system results in a hybrid neuro-fuzzy system.
u Hybridization: The main aim of the concept of hybridization is to overcome the weakness in one technique while applying it and bringing out the strength of the other technique to find a solution by combining them.
Neuro-fuzzy hybrid system (NFHS): - Proposed by J.S.R Jang (Jyh-Shing Roger Jang) in the early '90s.
u Neuro-Fuzzy hybridization is widely termed as Fuzzy Neural Network (FNN) or Neuro-Fuzzy System(NFS).
u NFHS is a learning mechanism that utilizes the training and learning algorithms from neural networks to find parameters of a fuzzy system.
u Neuro-Fuzzy Hybrid System is a combination of fuzzy system and neural network.
u The human-like reasoning style of fuzzy systems is incorporated by the use of
. Fuzzy sets
. Linguistic model
. along with IF-THEN fuzzy rules
Comparison of Fuzzy Systems with Neural Networks
A black box is a device, system or object which can be viewed in terms of its inputs and outputs without any knowledge of its internal workings.
Video Link: https://youtu.be/mV5vNaXypwc
The architecture of the neuro-fuzzy hybrid system
Ø The architecture is a three-layer feedforward neural network model.
Ø Second layer corresponds to the fuzzy rules.
Ø Third layer corresponds to the output variables.
Classification of the neuro-fuzzy hybrid system
u NFS are classified into the following two categories:
1. Cooperative NFSs
2. General Neuro-fuzzy hybrid systems
Cooperative NFS: Fuzzy system is governed by fuzzy IF-THEN rules.
u A fuzzy system is a set of fuzzy rules that convert inputs to outputs.
u A fuzzy rule is defined as a conditional statement in the form: IF x is A. THEN y is B. where x and y are linguistic variables; A and B are linguistic values determined by fuzzy sets on the universe of discourse X and Y, respectively.
u Linguistic variables: variable whose values are words or sentences rather than numbers. For E.g., Speed, Temp. and service.
u Linguistic value: Values or terms used to describe the linguistic variables. For E.g., for speed (slow, fast, faster), for temperature (cool, warm, hot), for service (good, poor, excellent.).
u IF temperature IS hot THEN Speed up fan.
u ANN attempt to learn parameters from the fuzzy system.
u FNN learns "fuzzy set" from the given "training data". This is done by fitting membership function with a neural network; fuzzy sets then being determined offline.
u A membership function (MF) is a curve that defines how each point in the input space is mapped to a membership value (or degree of membership) between 0 and 1.
General Neuro-fuzzy hybrid systems:
u Architecture of General Neuro-Fuzzy Hybrid Systems
u Advantage of general NFHS: there is no communication between NN and FS.
u The rule base of fuzzy system is assumed to be a NN.
u Fuzzy sets are regarded as weights.
u And the rules and input, output variables as Neurons.
u The choice to include or discard neurons can be made
In learning step.
u Fuzzy knowledge base is represented by neurons of NN.
Ø Membership functions expressing the linguistic terms Of the inference rule should be formulated for building Fuzzy controller.
Advantages of Neuro-fuzzy hybrid systems:
u It can handle any kind of information (numeric, linguistic, logical, etc.)
u It can manage imprecise, partial, vague or imperfect information.
u It has self-learning, self-organizing and self-tuning capabilities.
u It doesn’t need prior knowledge of relationships of data.
Areas of Applications for the use of Hybrid System:
u Engineering Design
u Stock market analysis and prediction
u Medical diagnosis
u Process control
u Credit card analysis
u Few cognitive simulations
No comments