The Rest of AIandSummaryCS472/CS473 – Fall 2005What have we done?• Learning– Theory: Generalization Error Bounds, Overfitting, CV– Algorithms: Decision Tree, KNN, Perceptron, NN, SVM• Acting– Theory: Markov Decision Processes– Reinforcement Learning: TD, ADP– Planning: Situation Calculus, STRIPS, Partial-Order Planning• Reasoning– Logic: FOL and Propositional– Reasoning: Resolution proofsSearch– Uninformed Search: DFS, BFS, IDS, Bi-Directional– Heuristic Search: Greedy, A*, IDS*– Local Search: Hill-Climbing, Sim. Annealing, Genetic– Constraint Satisfaction– Adversarial SearchWhat have we NOT done?• Applications– Acting: • Production planning in factories• Automated assembly– Learning:• Spam filtering• Credit card fraud detection• Market basket analysis– Reasoning:• Embedded diagnosis systems• Interactive help systems• Software verification• What is lacking? – Achieving everything together! A learning system that has knowledge and can plan and act under uncertainty.What have we NOT done?• Perception– Natural Language Understanding– Speech Recognition– Image Understanding–Etc.• Robotics– Speech Generation– Mechanics– Touch Sensing–Etc.Natural Language UnderstandingGoal: To create computational models of language in enough detail that you could write computer programs to perform various tasks involving natural language.Ultimate Goal: To be able to specify models that approach human performance in the linguistic tasks of reading, writing, hearing and speaking.Information RetrievalGoal: Choose from a set of documents the ones that are relevant to a query. Approach: Understanding via statistics.Example: Find me important WWW pages on SVMs.Harder Example: Find me all articles on leveraged buyouts involving more than 100 million dollars that were attempted but failed during 1986 and 1990.Information ExtractionGoal: Read a text and derive from it some assertions that can be put into a structured database. Approach: Restricted domain so that one can learn rules for individual assertions.Example:BOGOTA, 9 JAN 90 (EFE) – RICARDO ALFONSO CASTELLAR, MAYOR OF ACHI, IN THE NORTHERN DEPARTMENT OF BOLIVAR, WHO WAS KIDNAPPED ON 5 JANUARY, APPARENTLY BY ARMY OF NATIONAL LIBERATION (ELN) GUERRILLAS, WAS FOUND DEAD TODAY, ACCORDING TO AUTHORITIES. CASTELLAR WAS KIDNAPPED ON 5 JANUARY ON THE OUTSKIRTS OF ACHI, ABOUT 850 KM NORTH OF BOGOTA, BY A GROUP OF ARMED MEN , WHO FORCED HIM TO ACCOMPANY THEM TO AN UNDISCLOSED LOCATION. ÆDate: 05 JAN 90Location: COLOMBIA: BOLIVAR (DEPARTMENT): ACHI (TOWN)Type: KIDNAPPINGWeapon: *Victim: “RICARDO ALFONSO CASTELLAR” (“MAYOR OF ACHI”)Perpetrator: “GROUP OF ARMED MEN”Organization: “ARMY OF NATIONAL LIBERATION (ELN)”Machine TranslationGoal: Translate from one natural language to another.Examples:• The spirit is willing, but the flesh is weak.• The extension of the coverage of the health services to the underserved or not served population of the countries of the region was the central goal of the Ten-Year Plan and probably that of greater scope and transcendence.• Welcome to Chinese Restaurant. Please try your Nice chinese Food With chopsticks. the traditional and typical of Chinese glorious history and cultual. PRODUCT OF CHINA. Speech RecognitionGoal: Recognize spoken language and transcribe it to written language.Approach: Statistical Hidden Markov ModelsCombines:– What did it sound like?• Acoustic model of phonemes– How plausible is this sentence?• Read lots of text and gather probabilities of word n-gramsÎ What is the overall most likely sentence given the acoustic signal?Object DetectionExample: Face detectionApproach: machine learning (often SVM)Girosi et al.Object Recognition• Goal: Human pose recovery• Approach: Statistical graphical modelsHuttenlocher et al.Robots (and Softbots)Hondaaimaabac.comSonyOther AI Courses• COM S 475 Artificial Intelligence: Uncertainty and Multi-Agent Systems (Spring)• COM S 474 Introduction to Natural Language Processing (Fall)• COM S 478 Machine Learning (Spring)• COM S 430 Information Retrieval (Fall)• COM S 572 Heuristic Methods for Optimization (Fall)• COM S 578 Empirical Methods in Machine Learning and Data Mining(Fall)• COM S 630 Representing and Accessing Digital Information (Spring)• COM S 664 Machine Vision (Fall)• COM S 671 Introduction to Automated Reasoning (Fall)• COM S 672 Advanced Artificial Intelligence (Fall)• COM S 673 Integration of Artificial Intelligence & Operations Research (Fall)• COM S 674 Natural Language Processing (Spring)• COM S 676 Reasoning About Knowledge (Fall)• COM S 677 Reasoning About Uncertainty (Fall)• COM S 678 Advanced Topics in Machine Learning (Spring)Final and ProjectsFinal Exam:– Time: Monday, December 12, 9:00am - 11:30am, – Location: Upson Hall B17– Closed-book– Review Session: TBA– Additional reading: • Winston, Artificial Intelligence, 3rdedtition, Addison Wesley.• Mitchell, Machine Learning, McGraw Hill.Project Presentations– Time: Tuesday, December 13, 10:00am - 12:00am, 1:00pm – 3:00pm – Location: Olin Hall 165Have a good