Introduction to Artificial Intelligence

Transcription

1 Introduction to Artificial Intelligence Christian Jacob Department of Computer Science University of Calgary 1. What is Artificial Intelligence? How does the human brain work? What is intelligence? How do we emulate the human brain? How do we create intelligence? Who cares? Let s do some cool and useful stuff! 2. Basic Problem-Solving Strategies Basic search techniques Problem decompositon and AND/OR graphs Searching with problem-specific knowledge 2. Basic Problem-Solving Strategies Basic search techniques Problem decompositon and AND/OR graphs Searching with problem-specific knowledge 2.1 Basic Search Techniques Greedy search Gradient descent or ascent Stochastic search Simulated annealing Evolutionary search Depth-first vs. breadth-first search In search for the highest landmark... global maximum local maxima local maxima 1

2 2. Basic Problem-Solving Strategies Basic search techniques Problem decompositon and AND/OR graphs Searching with problem-specific knowledge 2.2 Problem Decomposition and AND/OR Graphs [Bratko, 2001] [Bratko, 2001] 2. Basic Problem-Solving Strategies Basic search techniques Problem decompositon and AND/OR graphs Searching with problem-specific knowledge 2.3 Searching with Problem-Specific Knowledge State h(n) A C B D E 80 G F H I A 366 B 374 C 329 D 2 4 E 253 F 178 G 193 H 98 I 0 A E C B h=253 h=329 h= Searching with Problem-Specific Knowledge (2) 3. Knowledge Representation, Reasoning, and Planning Knowledge representation Reasoning and planning Knowledge soup Anatoly Karpow and Gary Kasparow, 1986 [Kurzweil, 1990] [Newborn, 1997] 2

3 3. Knowledge Representation, Reasoning, and Planning Knowledge representation Reasoning and planning Knowledge soup 3.1 Knowledge Representation [Sowa, 2000] 3. Knowledge Representation, Reasoning, and Planning Knowledge representation Reasoning and planning Knowledge soup 3.2 Reasoning and Planning Start At(Home), Rents(Rogers, Video), Sells(Store, Bread) At(Home) At(Home) Go(Rogers) Go(Store) At(Rogers), Rents(Rogers, Video) Rent(Video) At(Store), Sells(Store, Bread) Buy(Bread) At(Home), Have(Video), Have(Bread) Finish 3. Knowledge Representation, Reasoning, and Planning Knowledge representation Reasoning and planning Knowledge soup 3.3 Knowledge Soup Vagueness, uncertainty, randomness, ignorance [Sowa, 2000] 3

4 4. Machine Learning and Pattern Recognition Fuzzy logic and fuzzy sets Artificial neural networks Single- and multilayer perceptrons Backpropagation networks Self-organizing feature maps Recurrent networks Neuro-fuzzy systems 4. Machine Learning and Pattern Recognition Fuzzy logic and fuzzy sets Artificial neural networks Single- and multilayer perceptrons Backpropagation networks Self-organizing feature maps Recurrent networks Neuro-fuzzy systems 4.1 Fuzziness What is hot? What is cold? What is young? What is old? 4. Machine Learning and Pattern Recognition Fuzzy logic and fuzzy sets Artificial neural networks Single- and multilayer perceptrons Backpropagation networks Self-organizing feature maps Recurrent networks Neuro-fuzzy systems [Sowa, 2000] 4.2 Artificial Neural Networks: Modeling the Brain 4.2 Artificial Neural Networks (2) Visual Cortex of a Cat Schematic Perceptron Feed-forward network [Stevens et al., 1988] [Kurzweil, 1990] [Spektrum, 1993] 4

5 Modeling the Brain? Optical Character Recognition Output Hidden [Spektrum, 1993] [Kurzweil, 1990] Input [Spektrum, 1993] 4. Machine Learning and Pattern Recognition Fuzzy logic and fuzzy sets Artificial neural networks Single- and multilayer perceptrons Backpropagation networks Self-organizing feature maps Recurrent networks Neuro-fuzzy systems 5. Evolutionary Computing 5. Evolutionary Computing 5.1 Evolution Strategies Evolution of a Jet Nozzle Evolutionary Engineering [Rechenberg, 1994] 5

6 5. Evolutionary Computing 5.2 Evolutionary Programming Input A 0 A0 output A 1 A 2 A 3 A 1 output A 2 output A 3 output Evolution of Finite State Automata [Jacob, 2001] 5.2 Evolutionary Programming 5. Evolutionary Computing Gen. 0 Gen. 0 Gen. 40 Gen. 40 Gen. 70 Gen. 70 Gen. 120 [Jacob, 2001] 5.3 Genetic Algorithms 5.3 Genetic Algorithms Genotype Phenotype Generation 0 Generation 11 Binary Vector {1,0,1,1,0,1,0,0,1,0,1,1} {0,1,1,1,1,0,0,1,0,0,0,1} {0,0,1,1,0,101,1,0,1,0,0} Decoding... {1,1,0,0,0,1,0,1,0,1,0,0}... {1,0,1,0,0,1,1,1,0,1,1,1} {0,0,1,1,0,1,1,1,0,1,0,0} {1,0,0,1,0,1,1,1,0,0,0,1} Interpretation Generation 10 Generation 30 [Jacob, 2001] 6

7 5. Evolutionary Computing 5.4 Genetic Programming LSystem LSystem Axiom _Axiom _LRules LRules PRED sprout LRule sproutindex SEQ BlankSequence Axiom LRules sprout[3] LRule LRule LEFT PRED SUCC RIGHT... [] F[_] Stack [] B[.38] Seq RL[20] F[4.9] leaf[0] Stack sprout[2] Seq bloom[1] LEFT sprout[3] SUCC Alternative _SEQ _STACK LRule _PRED RIGHT _SUCC Alternative F[1.4] YL[30] leaf[0] [] [] _sprout _stalk... _YL _SEQ advance] ] Gen. 1 advance], advance], Gen. 5 ifsensor[dust][stop]], advance]], ifsensor[dust][turnleft, again], nop]], advance], stop], nop], ifsensor[phero][turnleft], stop]]], turnright]]], nop]] ]] Gen. 16 stop], nop], ifsensor[phero][turnleft], advance]]], turnright]], ifsensor[dust][turnleft, again],nop]], turnleft, again], nop], ifsensor[phero][stop]]], advance]]]], nop]]]] Gen. 22 turnleft, again], turnright]], ifsensor[dust][turnleft, again],nop]], turnleft, again], nop], ifsensor[phero][stop]]], advance]]]], nop]] ]] Gen. 59 turnleft, turnleft, again], turnleft, again], turnright]], turnleft, again], nop], again], nop], nop]]]], advance]]]]]]], ifsensor[dust][turnleft, again]], ifsensor[dust][advance]]], ifsensor[dust][turnleft, again]], turnleft, again], nop], ifsensor[dust][stop], stop], turnleft]]], advance]]]], nop]]] ]]]] Fitness Generation turnleft, turnleft, again], turnleft, again], turnright]], turnleft, again], nop], again], nop], nop]]]], advance]]]]]]], ifsensor[dust][turnleft, again]], ifsensor[dust][ advance]]], ifsensor[dust][turnleft, again]], turnleft, again], nop], ifsensor[dust][stop], stop], turnleft]]], advance]]]], nop]]] ]]]] [Jacob, 2001] 5. Evolutionary Computing 5.5 Learning Classifier Systems A classifier system to emulate a frog. The frog reacts to objects it sees. Input: Moving On the Ground Output: Large Far Striped Flee! Pursue! _

8 5. Evolutionary Computing 5.6 Evolutionary Design: Objects Lifting Loads Scaffold [Funes and Pollack, 1999] 5.6 Evolutionary Design: Art 6. Swarm Intelligence and Complex Adaptive Systems Social Models Swarms and Emergent System Behaviour Immune System Computing Mutations Hölldobler & Wilson, Swarm Intelligence and Complex Adaptive Systems 6.1 Social Models: Competition and Cooperation Social Models Swarms and Emergent System Behaviour Immune System Computing [Nuridsany & Pérennou, 1996] [Ernst, 1998] 8

9 6. Swarm Intelligence and Complex Adaptive Systems 6.2 Swarms: The Ants Paradigm Social Models Swarms and Emergent System Behaviour Immune System Computing [Hölldobler & Wilson, 1990] 6.2 Swarms and Emergent System Behaviour 6. Swarm Intelligence and Complex Adaptive Systems Social Models Swarms and Emergent System Behaviour Immune System Computing 6.3 Immune System Computing 7. Robo Sapiens? Distinguishing self from non-self Intrusion detection in a LAN Seymour Papert [Hofmeyr and Forrest, 1999] [Kurzweil, 1990] LOGO Robot 9

10 7. Robo Sapiens? 7. Robo Sapiens? KISMET WABOT, theorgan Player Ichiro Kato, Waseda-University, Tokyo MIT [Menzel and D Aluisio, 2000] [Kurzweil, 1990] Artificial Intelligence in Action If we don t know how it works, then it s AI. When we find out how it works, it s not AI anymore References References (2) Bratko, I. (2001). PROLOG Programming for Artificial Intelligence. New York, Addison- Wesley. Rechenberg, I. (1994). Evolutionsstrategie 94. Stuttgart, Frommann-Holzboog. Sowa, J. F. (2000). Knowledge Representation. Pacific Grove, Brooks/Cole. Kurzweil, R. (1990). The Age of Intelligent Machines. Cambridge, MA, MIT Press. P. Menzel and F. D Aluisio (2000). Robo sapiens Evolution of a New Species. Cambridge, MA, MIT Press. Spektrum der Wissenschaft: Spezial. Gehirn und Geist. Heidelberg, Spektrum Akademischer Verlag,1993. Hofmeyr, S. and Forrest, S. (1999). Immunity by Design: An Artificial Immune System. In GECCO 99. Stevens, C. F., et al. (1988). Gehirn und Nervensystem. Heidelberg, Spektrum Akademischer Verlag. 10

11 References (3) References (4) Ernst, A. M., ed. (1998). Digest: Kooperation und Konkurrenz, Heidelberg, Spektrum Akademischer Verlag. Nuridsany, C., and Pérennou, M. (1996). Microcosmos: The Invisible World of Insects. New York, Stewart, Tabori & Chang. Newborn, M. (1997). Kasparov versus Deep Blue. Berlin, Springer-Verlag. Jacob, C. (2001). Illustrating Evolutionary Computation with Mathematica. San Francisco, Morgan Kaufmann. Hölldobler, B., and Wilson, E. O. (1990). The Ants. Cambridge, MA, Harvard University Press. Todd, S. and Latham, W. (1992). Evolutionary Art and Computers. London, Academic Press. Funes, P. and Pollack, J. (1999). Computer Evolution of Buildable Objects. In: P. Bentley (ed.). Evolutionary Design by Computers. San Francisco, Morgan Kaufmann. 11