1 / 24

Introduction to AI & AI Principles (Semester 1) REVISION WEEK 1 (2008/09)

Introduction to AI & AI Principles (Semester 1) REVISION WEEK 1 (2008/09). John Barnden Professor of Artificial Intelligence School of Computer Science University of Birmingham, UK. TODAY (Tuesday). Nature of exam (refining the info given in Week 11)

dylan-brown
Download Presentation

Introduction to AI & AI Principles (Semester 1) REVISION WEEK 1 (2008/09)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Introduction to AI &AI Principles (Semester 1)REVISION WEEK 1(2008/09) John Barnden Professor of Artificial Intelligence School of Computer Science University of Birmingham, UK

  2. TODAY (Tuesday) • Nature of exam (refining the info given in Week 11) • Review of material (extending the review in Week 11) • Questions • NB Office Hours: • Friday 1 May, 3:00-4:00 • Friday 8 May, 4:00-5:00

  3. Natureof the Examination

  4. Format of AI Principles Exam • Three hours long. • First half (1.5 hours): almost exactly the same as the Intro-AI exam (see next slide). • Second half: to be explained by Dean Petters in Revision Week 2. • Use of material: • First half can be done on the basis entirely of my material. • Second half can be done on the basis entirely of Dean’s material. • But you’re free to use his material in my half or mine in his half as appropriate.

  5. Format of AI Intro Exam • One and a half hours long. • Do 5 out of 6 questions. • Most question parts: broadly similar in style to exercises you did during Semester 1. • One question is essay-like and allows considerable latitude as to what aspects of AI you address and what material you bring to bear. • The rest are mostly on specific technical things, with a couple of free-wheeling aspects here and there.

  6. AI-Intro Material • My own lecture material, • with some exclusions (see Week 11 part of Slides page) • Answers / additional notes for Exercises. • Andrea Arcuri’s lecture on learning, with some exclusions. • Bullinaria slides (again with some exclusions): • Semantic Networks (and my own notes on these slides) • Production Systems (and my own notes on these slides) • Expert Systems • Textbook chapters (or chapter parts) in the Weekly Reading Assignments on module webpage.

  7. AI-Intro Material, contd • Don't be spooked by previous examinations, especially those from before 06-07!! • There have been a lot of changes. Also, quite a few since last year. • Knowledge of textbook chapters or chapter parts other than those I've listed ISN’Texpected. • Knowledge of Bullinaria slides other than those I point to from my list of weekly lecture slides ISN’T expected. • Knowledge of fine technical details in book chapters ISN’Texpected. (I’m only expecting the main concepts and overall grasp of main examples.) • But of course knowledge of all the above types could be helpful and impressive.

  8. REVIEWof the material

  9. Main Topics Covered • Representation and reasoning, in • logic • production systems • semantic networks. • What we need to represent: entities (incl. situations, feelings, …), properties, relationships, propositional structure, quantification, … • Planning (a type of reasoning). • Search. • Natural Language difficulties as illustration of why AI is difficult. • Knowledge and reasoning needed in natural language understanding and operating in practical scenarios such as Hot Drinks and Shopping Trip. • Learning.

  10. Main Detailed Techniques • Expressing information in logic. • Expressing information in semantic networks. • Applying production system rules (forwards or backwards, but fine detail only expected in forwards case). • Doing simple logical proofs. • Search (fine detail not expected for best-first and A*). • Search as applied to route-finding. • Search as applied to planning delivery of a drink.

  11. General Themes in AI • Why everyday AI is difficult. • Language processing, vision, planning, common-sense reasoning, etc. • “Intelligence” and its connection to “stupidity”. • What looks like stupidity is often the understandably-incorrect application of efficient heuristics (rules of thumb) without which we and our AI cousins would be in a mess. • Contd. ……

  12. General Themes in AI, contd. • Uncertain, vague, conflicting, missing, or diverse info. • Huge amounts of info, of varying relevance. • Hence: search, satisficing, graceful degradation, heuristics. • Context-sensitivity; incl. relativity to agents’ purposes (e.g., in vision and language interpretation). • Task variability, learning, adaptation, repair (e.g., of plans). • Declarative/procedural trade-off. • Goal-directedness (backwards chaining) in reasoning and search.

  13. A General Theme in AI • Uncertain, vague, conflicting, missing, diverse, extensive info: • Amply shown by Hot Drinks, Shopping Trip and Crime scenarios, • and by natural language examples. • Use of default rules and conflict resolution in PSs • Use of defaults and exceptions in SNs. • Contributes to need for search. Non-optimality (satisficing) in (some) search. • Use of heuristics in search. • Need for learning. • Graceful degradation in (e.g.) neural networks.

  14. A General Theme in AI • Search: • In planning (incl. route-finding, game-playing, …) • In deduction • In operation of Production Systems • In reasoning in Semantic Networks • In learning, particularly • genetic algorithms • automatically finding good weights for a neural network

  15. General Theme: Heuristics • PS rules that leave out details and complications, and that are at best DEFAULTS • Conflict resolution methods in PSs. • The information attached to actions in planning about what changes (or doesn’t change) is typically defeasible. • On what doesn’t change: see the Planning 1 chapter in Callan about the important frame problem. • In search in general: • Pruning • Action ordering in depth-first search • Evaluation functions in best-first search, incl. • Heuristic functions in A* search. • Choice of search strategy, incl. backwards vs. forwards.

  16. Rough Sequence of Topics • Introduction: • what AI is • why we do it • how it differs from ordinary CS • application areas • expert versus everyday AI.

  17. Topic Sequence contd: Challenge of AI • Introductory examples from language. • CAUTION CHILDREN • “John got to his front door but realized he didn’t have his key.” • Context-sensitivity of language; knowledge and reasoning needed. • Knowledge and reasoning needed in Hot Drinks, Shopping Trip and Crime scenarios. • Knowledge variety, uncertainty, vagueness, missing info, … • Vision and movement. • Context-sensitivity, purpose-sensitivity, ambiguity, …

  18. Sequence of Topics, contd. • Detailed planning of delivery of one drink. • Search, forwards versus backwards chaining, goal-directedness • Knowledge needed about preconditions and (non-)effects of actions • Search: general nature, example applications. • Introduction to logic representation. • Reasoning about a static situation using Production Systems. • CAUTION: different from planning, = reasoning about moving between different possible situations.

  19. Sequence of Topics, contd. • More on natural language difficulties. • vagueness • quantification subtleties • context-sensitivity • syntactic ambiguity, incl. PP attachment • some advanced topics: speech acts, mental states, metonymy, metaphor

  20. Sequence of Topics, contd. • Search detail (in route finding for Shopping Trip) • Different search strategies: depth-first, breadth-first, best-first, A* • Optimality or otherwise • Ordering and pruning heuristics • Evaluation/heuristic functions.

  21. Sequence of Topics, contd. • More on logic representation. • Logical deduction. • Inference rules in deduction versus production systems • Soundness • Fiddling around needed in deduction • Reduction of fiddling around by using Resolution • Reasoning by contradiction • Declarative/procedural trade-off. • Logical deduction versus using production systems. • Reasoning as search.

  22. Sequence of Topics, contd. • Representation and reasoning in Semantic Networks • Localization of info at nodes • Different types of link • Taxonomy (instances and subtypes) • Defaults and exceptions • Intersection search • More on Production Systems. • Rule instantiations • Conflict resolution • Expert Systems

  23. Sequence of Topics, contd. • Learning (Andrea Arcuri lecture in Week 9). • General characteristics • Neural Networks • Evolutionary Computation and Genetic Programming • Naïve Bayes Classifiers (not expected for exam)

  24. Questions

More Related