350 likes | 586 Views
G51IAI Introduction to Artificial Intelligence. Course Introduction. Andrew Parkes http://www.cs.nott.ac.uk/~ajp/. Joke Q. “What do you give a hurt lemon?” A. “Lemon aid” Hands up if you think the joke writer has some sense of humour? has a good sense of humour? is intelligent?
E N D
G51IAIIntroduction to Artificial Intelligence Course Introduction Andrew Parkes http://www.cs.nott.ac.uk/~ajp/
Joke • Q. “What do you give a hurt lemon?” • A. “Lemon aid” • Hands up if you think the joke writer • has some sense of humour? • has a good sense of humour? • is intelligent? • has a mental age of 5?, 10?, 15? • has probably been to too many lectures? • Joke writer is a computer program: JAPE
Joke Analysis and Production Engine (JAPE) • Author of JAPE: Kim Binsted • Ph.D. work in AI at Edinburgh (1996) • “an effort to combine her academic interest in artificial intelligence (AI) with her personal interest in improvisational comedy.” Daily Telegraph 1996 • More recently: STANDUP“System to Augment Non-speakers' Dialogue Using Puns” • “A computer system that generates simple word-play jokes could help disabled children develop better language skills, say UK researchers.” New Scientist 2006 • “Developed by Annalu Waller at Dundee University, UK, who has arranged for eight children with cerebral palsy to test the system.”
You are a caveman (or woman) • I travel back in time and bring you a Pentium IV PC and show you some of the things it is capable of doing. • Question : Would you, as a caveman, consider the computer to be intelligent? • Hands up if you think the computer is intelligent?
We are still in the present day, but we are at the end of this lecture course. • Just before I leave the final lecture, I peel away my face and reveal I am an android. As an android I have just delivered a 24 hour lecture course and answered all your questions. • Question : Would you, as a computer science student, consider me to be intelligent? • Hands up if you think I am intelligent?
Course Introduction • Course • Introduction to Artificial Intelligence • Lecturer • Andrew Parkes • http://www.cs.nott.ac.uk/~ajp/ • ajp ‘at’ cs.nott.ac.uk
Course Introduction • Web Page • http://www.cs.nott.ac.uk/~ajp/courses/g51iai/ • contains the self-study materials for this course • EMAIL • ajp ‘at’ cs.nott.ac.uk • Office Hours • Thurs 2-3 • or “by appointment” (email me) • Second / Third Year Projects • ??
Course Context G51IAI Introduction to AI G5BAIP Artificial Intelligence Programming G5BAIM Artificial Intelligence Methods D53DIA Designing Intelligent Agents, G53ASD Automated Scheduling, … Projects
Who Can Attend? (This year) • Introduction to AI • First/Second Year Option • Artificial Intelligence Methods • Second/Third Year Option
Course Introduction • Lectures • 15 (ish) • Lecture Times and Locations • Monday 14:00 (CTF-C33) • Wednesday 10:00 (New Business School. A25) • Assessment • 25% Coursework (CW1 5%, CW2 20%) • 75% examination (2 hours)
Aims of the Course • Define what we mean by AI (or at least give us a working definition for this course) • Know how to write “AI” programs that • Explore search spaces using both blind and heuristic search techniques • Implement Neural Networks (perceptron)
Aims of the Course • This is very much a practical course. Although we will touch upon the philosophical issues we will not dwell on this area of AI. • This course is more concerned with writing useful AI programs than discussing if a computer is intelligent or not.
Textbooks • Artificial Intelligence – A Modern Approach (AIMA) (Russell/Norvig)
Textbooks • Artificial Intelligence – A Modern Approach (AIMA) (Russell/Norvig) “Artificial Intelligence (AI) is a big field and this is a big book” (Preface to AIMA) 2nd Edition (2003) has 1081 pages You don’t need to learn them all You don’t even need to read them all
Textbooks • Artificial Intelligence – A Modern Approach (AIMA) (Russell/Norvig) Much of the material for this course is from this book. If you are going to buy a book, then get this one. “2nd edition 2003” is the one I use – though earlier editions are probably fine for the material in this course. BUT It can be quite an advanced book - sometimes more like a reference book than a textbook – so you might want to first consider “easier” books such as Cawsey
Textbooks • Artificial Intelligence – A Modern Approach (AIMA) (Russell/Norvig) • Relevant Chapters • Chap 1 : Introduction • Chap 3 : Solving Problems by Search • Chap 4.1 : Informed (Heuristic) Search Strategies • Sections 5.1 & 5.2 : Backtracking Search for CSPs • Chap 6 : Adversarial Search • Sections 11.1 & 11.2 : Planning • Section 20.5 : Neural Networks • Chap 26 : Philosophical Foundation • (You do NOT need everything from each of these chapters!!)
Useful Reading • The Essence of Artificial Intelligence (Cawsey) • Good lightweight introductory book • A good starting point for the course would be to read Chapter 4 on search • This book is at about the level of the course (in contrast, AIMA is generally more advanced than this course)
Textbooks • Artificial Intelligence (Rich/Knight) • This used to be the “standard” AI text book but AIMA is now taking its place – in my view
Textbooks • Artificial Intelligence (Winston) • As for Rich/Knight
Textbooks • Artificial intelligence : Structure and Strategies for Complex Problem Solving (Luger/Stubblefield)
Textbooks • Computational Intelligence (Poole/Mackworth/Goebel) • As good as AIMA but came out later
Useful Reading • Seven Methods for Transforming Corporate Data into Business Intelligence (Dahr/Stein)
Useful Reading • Artificial Intelligence and Computer Games (Richard Bartle)
Useful Reading • Computer Gamesmanship (David Levy)
Neural Networks • The Essence of Neural Networks (Callan)
Neural Networks • Neural Network (Davalo)
Neural Networks • Fundamentals of Neural networks (Fausett)
Course Structure Three distinct themes with crucial topics • Search • Tree Search : Breadth- and Depth-First Search • Graph Search • Uninformed : Uniform Cost Search • Informed Search : A* Search • State Space Search • Game Tree Search : minimax and alpha-beta • History & Philosophy • Turing Test & Chinese Room • Neural Nets: • Perceptrons & their limitations
Approximate Time Line Oct Nov Dec CW 1 (5%) CW 2 (20%) Exam (75%) Tree Search GraphSearch State Space Search Game Tree Search Part. Assign. Search Neural Nets History & Philosophy of AI
Proposed Lecture Schedule • See Web Site
Coursework Schedule • Coursework ONE (5%) • Is available now • It does not involve programming • Closing date is 1st November, • But with an automatic weeks extension! • No further extensions • Coursework TWO (20%) • Will be available before end of October • It does involve programming • Closing date is 30th November, • But with an automatic weeks extension! • No further extensions
Examinations • Examination Rubric • Is likely to be something such as • “You are expected to answer one of the two questions from section A and three of the five questions from section B.” • but not yet finalised.
Intro to AI: Quick Start to your Self Study • Have a look at the web materials • to make sure you can find and access them! • get an overview of the course • please report any problems • Find the textbooks in the library • Start (just start!) to read the search chapter of one of textbooks e.g. • Cawser : Chap 4 • Winston : Chap 4
Are these AI?? Boids