Spring, 2005 CSE391 – Lecture 1 1 Introduction to Artificial Intelligence Martha Palmer CSE391 Spring, 2005
CSE391 – Lecture 1 2 Robots
Spring, 2005 CSE391 – Lecture 1 3 Outline What is “artificial” intelligence? Building intelligent agents Course Outline
Spring, 2005 CSE391 – Lecture 1 4 What does “artificial” intelligence mean? Programming a computer to successfully perform tasks that are thought to require intelligence –Playing chess –Proving theorems –Translating Russian into English –Walking across a room –Recognizing a familiar face –Understanding directions
Spring, 2005 CSE391 – Lecture 1 5 Computers are Good at: –Number crunching –Storing information –Airline scheduling –Transmitting data –Structured data bases –Graphics Bad at: –Writing poetry –Composing music –Understanding speech –Driving cars –Enjoying peaches –Learning new things
Spring, 2005 CSE391 – Lecture 1 6 Artificial Intelligence Human Intelligence! How do we build intelligent machines?
Spring, 2005 CSE391 – Lecture 1 7 Building Intelligent Agents: First Challenge Create a representation of the world (the task) in terms computers can deal with –Numbers? –Strings If then else statements features Let’s assume everything about the task can be represented – we have complete knowledge
Spring, 2005 CSE391 – Lecture 1 8 Building Intellligent Agents: Second Challenge Extend our programs to handle situations where knowledge isn’t complete, i.e., where there is uncertainty
Spring, 2005 CSE391 – Lecture 1 9 Building Intelligent Agents Agent prior knowledge actions experience goals/values observations
Spring, 2005 CSE391 – Lecture 1 10 Intelligent Agent skills include: ReasoningSearchMachine Learning Representation of the World Symbols (Logic, Numbers) Vision Processing Planning Robotics Natural Language Understanding
Spring, 2005 CSE391 – Lecture 1 11 Intelligent Agents Agent prior knowledge actions experience goals/values observations
Spring, 2005 CSE391 – Lecture 1 12 Applications Medical diagnosis Interactive Fiction Machine Translation Autonomous vehicles Multilingual Google Deep Blue
Spring, 2005 CSE391 – Lecture 1 13 Course Outline – First Half (7 weeks ) Intelligent Agents (1 week) Search (3 weeks) Blind search: Breadth-first search, Depth-first search, Iterative Deepening Using heuristics: A* algorithm, Iterative deepening A* Search implementation (breadth-first, depth- first, IDA, A*) Research Paper Topics: games, parser search spaces, hill-climbing, simulated annealing
Spring, 2005 CSE391 – Lecture 1 14 Course Outline - First half, cont. –Using Logic for Knowledge Representation and Reasoning (2 weeks) Reasoning with Propositional Logic Research paper topics: resolution theorem proving, 4 color map, proving program correctness, fuzzy logic, modal logic
Spring, 2005 CSE391 – Lecture 1 15 Course Outline - Second half (Six weeks) Dealing with the real world and uncertainty –Natural Language Processing (2 weeks) Syntax and semantics: Running a parser Research Paper topics: Speech recognition, Information Extraction, Machine Translation, Eliza –Machine Learning (2 weeks) Decision Trees: WSD Implementation using WordNet Research Paper Topics: neural networks, connectionism, applications, ontologies (CYC, Semantic Web, etc.), –Probabilistic Reasoning/Bayesian nets (2 weeks) Bayesian Net exercises and implementation Research Paper Topics: Applications, expert systems, medical applications, decision support systems,
Spring, 2005 CSE391 – Lecture 1 16 Course Outline – last bit (Two weeks) Project presentations (2 weeks)
Spring, 2005 CSE391 – Lecture 1 17 Course Structure 35% Homeworks 25% Research project – paper or program 15% Major exam – Search, KRR, 15% Major exam – NLP, Dec. Trees, Bayes, 10% Class participation