Top Banner

Click here to load reader

Artificial intelligence & expert systems

Jul 05, 2015

ReportDownload

Education

yoga-priya

A brief introduction to Artificial Intelligence & Expert Systems

  • 1. ARTIFICIALINTELLIGENCE

2. Different perspectives DefinitionIntelligence artificial intelligence is making machines"intelligent" -- acting as we would expect peopleto act.Research "artificial intelligence is the study of how to makecomputers do things which, at the moment,people do better"Business AI is a set of very powerful tools, andmethodologies for using those tools to solvebusiness problems.Programming AI includes the study of symbolic programming,problem solving, and search. 3. TypesExpert systems obtained information from experts areconverted in to codes based on thumb rulesto be followedNeural networks Animals and people naturally distinguishmany kinds of complex patterns, such as thesound of a bird or the shape of a face. Somekinds of computer programs "learn" things inways that mimic the behaviour of biologicalnerve cells.Motion controllers: Some researchers study how people andanimals move about and manipulate objectsto improve the motion of robots andmachinesGenetic algorithms : Genetic algorithms take a competitive andrepetitive approach to problem-solving. 4. Tasks ApplicationFormal Tasks mathematics, gamesMundane tasks perception, robotics, natural language,common sense reasoning)Expert tasks financial analysis, medical diagnostics,engineering, scientific analysis, and otherareas 5. Machine Learning!Machine learning is a scientific discipline concerned with the designand development of algorithms that allow machines to mimic humanintelligence. 6. How Does Artificial Intelligencelearn? Its a program that learn by makingmistakes so that mistakes are notrepeated.Failure drivenlearning Here a teacher has to teach. Butcommunication is the problem. Hencecode languages are usedLearning bybeing told Here it does not work toward goal butjust keeps on acquiring information tolearn so data base keeps on increasing.Learning byexploration 7. Turing Test It is a test of machines ability to demonstrateintelligence 8. Human Intelligence v/s ArtificialintelligenceHuman Intelligence Artificial Intelligence Humans are fallible They have limitedknowledge bases Information processing ofserial nature proceedvery slowly in the brain ascompared to computers Humans are unable toretain large amounts ofdata in memory. No common sense Cannot readily deal withmixed knowledge May have highdevelopment costs Raise legal and ethicalconcerns 9. Human Intelligence v/s Artificialintelligence Intuition, Commonsense, Judgment,Creativity, Beliefs etc The ability todemonstrate theirintelligence bycommunicatingeffectively Plausible Reasoningand Critical thinking Ability to simulatehuman behavior andcognitive processes Capture and preservehuman expertise Fast Response. Theability to comprehendlarge amounts of dataquickly 10. Artificial Intelligence VS ConventionalComputingArtificial IntelligenceConventionalComputing AI software uses thetechniques of searchand pattern matching Programmers designAI software to give thecomputer only theproblem, not the stepsnecessary to solve it Conventionalcomputer softwarefollow a logical seriesof steps to reach aconclusion Computerprogrammersoriginally designedsoftware thataccomplished tasksby completingalgorithms 11. Expert Systems An expert system is software that attempts toreproduce the performance of one or morehuman experts, most commonly in a specificproblem domain. Ex, Play chessHelp in financial decisionsConfiguration of computers etc 12. Major Components Knowledge base - a declarative representation of theexpertise, often in IF THEN rules Working storage - the data which is specific to a problembeing solved Inference engine - the code at the core of the system Derives recommendations from the knowledge base andproblem-specific data in working storage User interface - the code that controls the dialog between theuser and the system 13. KnowledgeBaseUserInterfaceInferenceEngineWorkingStorageDomainExpertKnowledgeEngineerSystemEngineerUSERExpertiseEncodedExpert 14. Terms ImportanceDomain expert The development of accounting software requiresknowledge in two different domains, namely accounting andsoftwareKnowledge engineer KE is an engineering discipline that involvesintegrating knowledge into computer systems in order tosolve complex problems normally requiring a high levelof human expertise.Systems engineering It is an interdisciplinary field of engineering that focuses onhow to design and manage complex engineering projectsover their life cycles. Systems engineering deals with work-processes,optimization methods, and riskmanagement tools in such projects.User User will be consulting with the system to get advice whichwould have been provided by the expert 15. Characteristics Of ExpertSystems The Highest level of expertise Right on time reaction Accepting the incorrect reasoning Good reliability Easily understood Flexible Symbolic reasoning Heuristic reasoning Making mistakes Expanding with tolerable difficulties 16. Advantages v/s DisadvantagesAdvantages Disadvantages Consistent answers forrepetitive decisions,processes and tasks Holds and maintainssignificant levels ofinformation Encouragesorganizations to clarifythe logic of theirdecision-making Never "forgets" to ask aquestion, as a humanmight Lacks common sense Cannot make creativeresponses as humanexpert Domain experts notalways able to explaintheir logic andreasoning Errors may occur in theknowledge base Cannot adapt tochanging environments Costly to develop Legal & ethical dilemma Difficult to use

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.