Applied Artificial Intelligence NOTES (1).pptx

Notes

1. APPLIED ARTIFICICIAL INTELLIGENCE

2. Definition Artificial intelligence is the effort to develop computer based systems designed to behave like humans, with the ability to learn languages, accomplish physical tasks, use perceptual apparatus and emulate human expertise and decision making

3. Attributes of an Intelligent Behaviour Think and reason Use reason to solve problems Learn or understand from experience Acquire and apply knowledge Exhibit creativity & imagination Deal with complex or perplexing situations Respond quickly & successfully to new situations Recognize the relative importance of elements in a situation Handle ambiguous, incomplete or erroneous information

4. Major Application Areas of A.I Expert Systems Learning Systems Fuzzy logic Genetic algorithms Neural Networks Intelligent Networks Visual perception Tactility (tangible ,physical) Dexterity (Skill in handling manipulation Locomotion Navigation Natural Languages Speech recognition Multi_sensory interface Virtual reality

5. Contribution of AI in Business To preserve expertise that might be lost through the retirement, resignation, or death of an acknowledged expert. To store information in an active form to create an organizational knowledge base that many employees can examine, much like an electronic textbook or manual, so that others may learn rules of thumb not found in textbooks. AI is of interest to business for the following reasons

6. To create a mechanism that is not subject to human feelings such as fatigue & worry. This may be especially useful when jobs may be environmentally, physically, or mentally dangerous to humans. These systems also maybe useful advisors in times of crisis. To eliminate routine & unsatisfying jobs held by people. To enhance the organization’s knowledge base by suggesting solutions to specific problems that are too massive & complex to be analysed by human beings in a short period of time.

7. Expert Systems An expert system is a knowledge intensive computer program that captures the expertise of a human in a limited domain of knowledge

8. Components of an Expert System User User Interface program Knowledge Base Knowledge acquisition Program Expert and/or Knowledge Engineer Inference Engine Program (Workstation) Exp. Sys. development Expert sys s/w

9. a) Knowledge Base Contains Facts about a specific subject area (e.g Jane is sick) Heuristics (rules of thumb) that express the reasoning procedures of an expert on the subject.(e.g IF Jane is sick, THEN she needs treatment)

10. b) Knowledge Acquisition Programs These are software tools for knowledge base development (not part of an expert system ) used as expert system shells, for developing expert systems.

11. c) The Inference Engine Is the software that executes the reasoning I.e processes the knowledge such as rules & facts related to a specific problem. It then makes associations and inferences resulting in recommended courses of action for a user.

12. d) User Interface Programs: These are for communicating with end users and includes an explanation program to explain the reasoning process to a user if requested.

13. Applications of Expert Systems Loan Applications: This is used by financial institutions. The user enters certain key facts into the system such as the loan applicant’s name, address, their income, monthly outgoings, and details of other loans. The system will then. Check the facts given against its database to see whether the applicant as a good previous credit record.

14. Perform calculations to see whether the applicant can afford to repay the loan. Make judgment as to what extent the loan applicant fits the lender’s profile of a good risk (based on the lender’s previous experience). Suggest a decision.

15. Legal advice. Tax advice. Forecasting: e.g Forecasting of economic of financial developments or of market and customer behaviour. Surveillance: e.g of the number of customers entering a supermarket, to decide what shelves need restocking and when more checkouts/branches need to be opened , or of machines in a factory, to determine when they need maintenance.

16. Diagnostic Systems: To identify causes of problems e.g in production control in a factory, or in healthcare. Education and Training : Expert systems can be used in diagnosing a student’s or worker’s weaknesses and providing or recommending extra instruction as appropriate .

17. Requirements for Building E/S An expert system, can be dvpt by one or more domain experts and knowledge engineers in a team. OR. Experts skilled in the use of expert system shell could develop their own expert system.

18. Requirements Use of an expert system shell: Is a low cost software package that consists of an expert system without its knowledge base (kernel). They allow trained users to develop the knowledge base for a specific expert system application on PCs, e.g a shell uses a spreadsheet format to help end users develop IF- THEN – ELSE rules automatically.

19. A Knowledge Engineer. Is a specialist who elicits information and expertise from other professionals and translates it into a set of rules or frames for an expert system . Expert (s). Is an individual with the expertise desired.

20. Advantages of Expert Systems AI and expertise is permanent, whereas human experts may leave the business. AI is easily copied. AI is consistent , whereas human experts and decision makers may not be. AI can be documented.The reasoning behind an expert recommendation produced by a computer will be recorded. Depending on the task the computer may be much faster than the human being.

21. Disadvantages of Expert Systems Systems are expensive. The technology is still relatively new. Systems will probably need extensive testing and debugging. People are naturally more creative (deal with unstructured data). Systems have very narrow focus.

22. Moral issues Fears of the experts losing their expertise beyond their control. The potential for royalty issues. Redundancy fears.

23. CASE BASED REASONING This is the artificial intelligence technology that represents organisation’s collective knowledge and expertise build up over the years as a database of cases and solutions. When a user enters a new case, the system searches for the stored cases with problems characteristics similar to the new one, finds the closest fit, and applies the solutions of the old case to the new case. Successful solutions are tagged to the new case & both are stored together with other cases in the knowledge base.

24. How CBR works User describes the problem Sys searches d/base for similar cases Sys asks user add’nal qtns to narrow the search Sys finds closest fit & retrieves soln Sys modifies the soln to better fit the problem Successful ? Sys stores problem & successful sol’n in the d/base Case Database 1 2 3 4 5 6 Yes No

25. Learning Systems Adaptive learning systems can modify their behaviours based on information they acquire as they operate e.g chess playing systems.

26. Genetic Algorithms ( adaptive Computation) Is problem solving methods that promote the evolution of solutions to specified problems conceptually using Darwinian process of evolution/survival for the fittest. (I.e the model of living organisms adapting to their environment.). They are programmed to work the way populations solve problems I.e by changing and reorganizing their component parts using processes such as reproduction, mutation & natural selection.

27. Thus genetic algorithms promote the evolution of solutions to particular problems, controlling the generation, variations, adaptation & selection of possible solutions using genetically based processes. As solns alter & combine, the worst ones are discarded and the better ones survive to go on to produce even better solutions.

28. Application Is suitable for very dynamic & complex situations involving randomizing & mathematical functions that have hundreds of variables or formulas. Solutions to certain types of problems in areas of optimization, product design etc. optimization: *Min. costs & max Profits. *efficient scheduling. * Use of resources. Commercial applications e.g design of jet turbine aircraft engines (involves like 100 variables & 50 constraints equations).

29. Virtual Reality Class Presentation