EE04 804(B) Soft Computing Ver. 1.2 Class 1. Introduction February 21st,2012 Sasidharan Sreedharan www.sasidharan.webs.com 1
Syllabus Artificial Intelligence Systems- Neural Networks, fuzzy logic, genetic algorithms, Artificial neural networks: Biological neural networks, model of an artificial neuron, Activation functions, architectures, characteristics- learning methods, brief history of ANN research- Early ANN architectures (basics only)- McCulloh & Pitts model, Perceptron, ADALINE, MADALINE
Objective To acquaint the students with important soft computing methodologies neural networks, fuzzy logic, genetic algorithms, and genetic programming. 3
What is meant by soft computing? Definition: Soft computing refers to a consortium of computational methodologies by including components such as Fuzzy Logic, Neural Networks, Genetic Algorithms etc in Artificial Intelligence platform to apply the acquired information to new conditions. Text Book: Neural Networks, Fuzzy Logic and Genetic Algorithms Synthesis and Applications S Rajasekaran and Vijayalakshmi Pai 4
What is meant by soft computing? Soft computing deals with imprecision, uncertainty, partial truth, and approximation to achieve practicability, robustness and low solution cost. Components of soft computing include: Neural networks (NN) Fuzzy systems (FS) Evolutionary computation (EC), including: Evolutionary algorithms Harmony search Swarm intelligence Ideas about probability including: Bayesian network Chaos theory Perceptron Soft computing techniques resemble biological processes more closely than traditional techniques, which are largely based on formal logical systems, such as sentential logic and predicate logic, or rely heavily on computer-aided numerical analysis (as in finite element analysis). Unlike hard computing schemes, soft computing techniques exploit the given tolerance of imprecision, partial truth, and uncertainty for a particular problem. Inductive reasoning plays a larger role in soft computing than in hard computing. 5
Soft computing (SC) Objective: Mimic human reasoning Main constituents: Neural networks Fuzzy systems Evolutionary Algorithms Genetic Algorithm 6
Constituents of SC Fuzzy systems => imprecision Neural networks => learning Evolutionary computing => optimization Over 24 000 publications today 7
Soft Computing Verses Hard Computing The term soft computing was introduced by Lotfi A Zadeh of the university of California, Berkeley, USA Soft Computing differs from hard computing (conventional computing) in its tolerance to imprecision, uncertainty and partial truth. Hard computing methods are predominantly based on mathematical approaches and demand a high degree of precision and accuracy. In engineering problems, the input parameter cannot be determined with high degree of precision. The role model for soft computing is human mind, biological systems. A powerful means for obtaining solutions to problems quickly. The guiding principle of soft computing is to accept the tolerance for imprecision, uncertainty and partial truth to achieve tractability, robustness and low cost solution. It is part of intelligent system. 8
Intelligent systems Intelligence: System must perform meaningful operations. Interprets information. Comprehends the relations between the phenomena or objects. Applies the acquired information to new conditions. 9
Advantages of SC Models are based on human reasoning. Models can be - simple - comprehensible - fast when computing - good in practice 10
Integration of soft computing technologies 11
Neural networks Simplified model of biological nervous system analogous to human brain with large number of neurons. Learns by example (Supervised learning and unsupervised learning) Once trained, the network can be put to effective use in solving unknown or untrained instances of the problem. Different architectures such as single layer feed and multi layer network. Can be applied to problems in pattern recognition, image processing, data compression, forecasting, optimization etc. 12
Neural networks (NN, 1940's) Neural networks offer a powerful method to explore, classify, and identify patterns in data. In p u ts N eurons (1 layer) O utputs Website of Matlab Neuron: y= w i x i 13
Fuzzy Logic Fuzzy set theory proposed by Lotfi A zadeh. Generalization of classical set theory. Fuzzy logic representations founded on Fuzzy set theory try to capture the way humans represent and reason with real world knowledge in the face of uncertainty. Wide applications in consumer electronics. Fuzzy Logic Washing Machine Fuzzy Logic Rice Cooker 14
Fuzzy Logic Deal with imprecise entities in automated environments (computer environments) Base on fuzzy set theory. Most applications in control and decision making Matlab's Fuzzy Logic Toolbox Omron s fuzzy processor 15
Y Model construction (SC/fuzzy) - Approximate values - Rules only describe typical cases (no rule for each input). => Small rule bases. - A group of rules are partially fired simultaneously. If x 0, then y 1 If x 5, then y 0.5 If x 10, then y 0 1,2 1 0,8 0,6 0,4 0,2 0 0 2 4 6 8 10 12 X 16
Genetic Algorithms Developed in 1970 by John Holland. Random search which mimic some of the processes of natural evolution. Based on a qualifying function termed as fitness function.(fitness means figure of merit) Genetic operators such as reproduction, cross over, mutation etc are used. Used for optimization applications 17
SC applications: control Heavy industry (Matsushita, Siemens,Stora-Enso) Home appliances (Canon, Sony, Goldstar, Siemens) Automobiles (Nissan, Mitsubishi, Daimler-Chrysler, BMW, Volkswagen) Spacecrafts (NASA) 18
SC applications: business supplier evaluation for sample testing, customer targeting, sequencing, scheduling, optimizing R&D projects, knowledge-based prognosis, fuzzy data analysis hospital stay prediction, TV commercial slot evaluation, address matching, fuzzy cluster analysis, sales prognosis for mail order house, multi-criteria optimization etc. (source: FuzzyTech) 19
SC applications: robotics Joseph F. Engelberger We are proud to announce that the HelpMate Robotic Courier has been acquired by Pyxis Corporation. Entertainment robot AIBO Fukuda s lab 20
SC applications: others Statistics Social sciences Behavioural sciences Biology Medicine 21
SC and future SC and conventional methods should be used in combination. 22
References 1. J. Bezdek & S. Pal, Fuzzy models for pattern recognition (IEEE Press, New York, 1992). 2. L. Zadeh, Fuzzy logic = Computing with words, IEEE Transactions on Fuzzy Systems, vol. 2, pp. 103-111, 1996. 3. L. Zadeh, From Computing with Numbers to Computing with Words -- From Manipulation of Measurements to Manipulation of Perceptions, IEEE Transactions on Circuits and Systems, 45, 1999, 105-119. 4. L. Zadeh, Toward a theory of fuzzy information granulation and its centrality in human reasoning and fuzzy logic, Fuzzy Sets and Systems 90/2 (1997) 111-127. 5. H.-J. Zimmermann, Fuzzy set theory and its applications (Kluwer, Dordrecht, 1991). 23