Expert Systems Concepts

صفحه 1:


صفحه 2:
Expert Systems Concepts 

صفحه 3:
Artificial Intelligence A field of study in computer science that pursues the goal of making a computer reason in a manner similar to humans. (pene eed exe Cee ee eae ole ‏ا‎ 

صفحه 4:
Human expert Vs. Expert system ات ا ا ل ةا ل اناه كت انال ‎reasonable to use it instead of a human expert?‏ Yes, because: Human expert may make mistakes, too. The expert system has the following advantages: سكير روود ‎Factor Human expert‏ ‎Bra‏ 0 993 معط رنه ‎teectl‏ 0 ‎econ 0000 availability‏ و 0 7 ‎ont‏ تون ممصمو كوم حماعه وآقتاهن: ‎fons:‏ affordable 

صفحه 5:
Fields of usage: Controlj 5 ‎Design +1,‏ تشخیصب پملیبا عیب1(18010516 ارا | ‏تفسير 172661701612161012 نظیتبر عملکرد 011060] 1۷۲09 برنامه ييزي و متصطة21 ‏پیش ینی05 ا نت ‎Prescription ۳ ‏انتخاب‌طم 56160 ‎Simulation <j +s 

صفحه 6:
Expert system structure: edge Bi User Engine ‏لضا‎ © Main difference of an Expert system and ordinary control programs: Knowledge is separate from Control Sa leet eee ge ene) دانشى وجود ندارد. آبن برخلاف برنامه نويسى معمولى است كه همه حا و ‎Cee‏ و يت 399 

صفحه 7:
Example: ‏اس سسکا‎ ۹ ‎ee od‏ ۱] مجادبد سات ا ‎١‏ جا بيب واحاموم ‎Ce‏ ‎er)‏ بمصفصا ذا 90:00 بجا سجاه حا قحا حا دحا ل ل ل ‎User: True (The car does not start)‏ ‎Nueva emit nacre tcc Ment CyO}‏ الت (فكله إفنص) 107 > موفغله؟ بمعققط مط 15 تسعادرد 1 ‎User: True (The battery voltage is < 10)‏ ‎> The fault is a bad battery (Rule2) ‎00000 ‎Dee ee ete cree eo ۱ et aC ‎ ‎ ‎ ‎ ‎ ‎ ‎ 

صفحه 8:
Explanation Bee darren ieee tig ee re ‏از‎ ‏0خ‎ = aaa ee ee ‏سيو ع‎ ۳ a he) 19 ‏ا‎ ‎eae 

صفحه 9:
Heuristic reasoning: A kind of inexact oning 

صفحه 10:
وت ات روک ا ل ۱ 200 ل يسا ۳ 0 ser Ee 8 

جلسه اول Expert Systems Concepts Artificial Intelligence A field of study in computer science that pursues the goal of making a computer reason in a manner similar to humans. Expert system: A computer program designed to model the problem-solving ability of a human expert. Human expert Vs. Expert system The expert system may make mistakes. So, is it reasonable to use it instead of a human expert? Yes, because: Human expert may make mistakes, too. The expert system has the following advantages: Factor Human expert Expert system Time of availability Geographic Safety Performance Cost workday local irreplaceable variable high always anywhere availability replaceable consistent, usually faster affordable Fields of usage: Controlک نترل Design ط راحي Diagnosisت شخيصب يماريي ا ع يب Instruction آموزش Interpretation ت فسير Monitoring ظارت ر ع ملکرد ب ن Planning ب رنام ه ريزي Predictionيش يني ب پ Prescription ت جويز Selectionان تخاب Simulation ش بيه س ازي Expert system structure: ‏User ‏Inference ‏Engine ‏Knowledge ‏Base ‏Working ‏Memory ‏Main difference of an Expert system and ordinary ‏control programs: ‏Knowledge is separate from Control (دانش سيستم در جايي مجزا از موتور استنتاج ذخيره شده و در کد مربوط به خود موتور استنتاج دانشي وجود ندارد .اين برخالف برنامه نويسي معمولي است که همه حاالت را در قالب If- elseهاي تودرتو و پيچيده در خود برنامه اصلي مي نويسيم) . Example: Knowledge Base (Rules) Rule1: If the car will not start Then the problem may be in the electrical system Rule2: If the problem may be in the electrical system AND the battery voltage is <10 Then the fault is a bad battery  Expert system: Does the car not start? (True/ false)  User: True (The car does not start)   The problem may be in the electrical system (Rule1)  Expert system: Is the battery voltage < 10? (True/ false)  User: True (The battery voltage is < 10)   The fault is a bad battery (Rule2) Working Memory (Facts) The car does not start The problem may be in the electrical system The battery voltage is < 10 ت وضيح ‏Explanation سيستمهاي خبره بعضا قابليت توضيح دادن منطق خود را نيز دارند .مثال اگر نتيجه اي را به کاربر اعالم کند و کاربر بخواهد بداند چظور به اين نتيجه رسيده بايستي مسيري را که طي کرده (از چه factهايي طبق چه قوانيني به ترتيب به نتيجه رسيده) براي کاربر شرح دهد. ‏Expert system: The fault is a bad battery ?User: How يا اگر براي روشن شدن برخي factهاي نامعلوم از کاربر سوال مي پرسد ،اگر کاربر خواست بداند که چرا اين سوال را مي پرسد ،بايد دليلش را بگويد .مثال در مثال زير بايد به کاربر بگويد که اگر جواب تو trueباشد آنگاه احتماال اشکال از سيستم برق است .آنگاه اگر ولتاژ باطري کمتر از 10 باشد معلوم مي شود که باطري خراب است .يعني بر خالف مثال قبل که روال طي شده را توضيح مي داد در اينجا روالي را که مي خواهد طي کند توضيح مي دهد. )Expert system: Does the car not start? (True/ false ?User: Why Heuristic reasoning: A kind of inexact reasoning  Heuristic knowledge :     I always check the electrical system first People rarely catch a cold during summer If I suspect cancer, I always check the family history. Old pipes often v ibrate under low flow rates  Heuristic re asoning (based on a heuristic knowledge):    If pipes are vibrating AND the pipes are old Then suspect low flow rate  Inexact knowle dge:    Bob's food are usually good If EKG test results aren't av ailable, but the patient is suffering chest pains, I might suspect a heart pr oblem Add a little oil to a motor that is running hot.  Inexact information (facts ):    I may buy food from Bob. I don't have EKG test r esults. The motor is running hot. Knowledge engineering: ش ( )م هندسيدان The process of building an expert system.  Involved persons to an Expert System:  Domain expert (.نيم... ميک. استفاده. خبره.يستم..حيس. از دانشاو در طرا.ه...ظر ک.. مورد ن. در زمينه.رد خبره...ف.)  Knowledge engineer (. خبره.يستم..ح س.)طرا  End user (. خبره.يستم..اربر س...)ک