Do not start until you are given the go-ahead signal

Similar documents
Lecture 10: Reinforcement Learning

Chapter 2. Intelligent Agents. Outline. Agents and environments. Rationality. PEAS (Performance measure, Environment, Actuators, Sensors)

Agents and environments. Intelligent Agents. Reminders. Vacuum-cleaner world. Outline. A vacuum-cleaner agent. Chapter 2 Actuators

Intelligent Agents. Chapter 2. Chapter 2 1

Decision Analysis. Decision-Making Problem. Decision Analysis. Part 1 Decision Analysis and Decision Tables. Decision Analysis, Part 1

Axiom 2013 Team Description Paper

Self Study Report Computer Science

Introduction to Simulation

Firms and Markets Saturdays Summer I 2014

AMULTIAGENT system [1] can be defined as a group of

Discriminative Learning of Beam-Search Heuristics for Planning

Reinforcement Learning by Comparing Immediate Reward

Chapter 2 Rule Learning in a Nutshell

Evolution of Collective Commitment during Teamwork

THE EFFECTS OF TEACHING THE 7 KEYS OF COMPREHENSION ON COMPREHENSION DEBRA HENGGELER. Submitted to. The Educational Leadership Faculty

Language properties and Grammar of Parallel and Series Parallel Languages

Spring 2016 Stony Brook University Instructor: Dr. Paul Fodor

A Version Space Approach to Learning Context-free Grammars

Learning and Transferring Relational Instance-Based Policies

Functional Skills Mathematics Level 2 assessment

Visual CP Representation of Knowledge

Version Space. Term 2012/2013 LSI - FIB. Javier Béjar cbea (LSI - FIB) Version Space Term 2012/ / 18

Learning Methods in Multilingual Speech Recognition

How to make successful presentations in English Part 2

RETURNING TEACHER REQUIRED TRAINING MODULE YE TRANSCRIPT

MYCIN. The MYCIN Task

While you are waiting... socrative.com, room number SIMLANG2016

Liquid Narrative Group Technical Report Number

Rule-based Expert Systems

An OO Framework for building Intelligence and Learning properties in Software Agents

CS Machine Learning

Designing A Computer Opponent for Wargames: Integrating Planning, Knowledge Acquisition and Learning in WARGLES

ENEE 302h: Digital Electronics, Fall 2005 Prof. Bruce Jacob

University of Groningen. Systemen, planning, netwerken Bosman, Aart

A Grammar for Battle Management Language

The Task. A Guide for Tutors in the Rutgers Writing Centers Written and edited by Michael Goeller and Karen Kalteissen

Level 6. Higher Education Funding Council for England (HEFCE) Fee for 2017/18 is 9,250*

Learning goal-oriented strategies in problem solving

AP Statistics Summer Assignment 17-18

Name Class Date. Graphing Proportional Relationships

ACTL5103 Stochastic Modelling For Actuaries. Course Outline Semester 2, 2014

Introduction and Motivation

Radius STEM Readiness TM

Knowledge-Based - Systems

Ricochet Robots - A Case Study for Human Complex Problem Solving

BMBF Project ROBUKOM: Robust Communication Networks

Probability and Game Theory Course Syllabus

Can Money Buy Happiness? EPISODE # 605

ISFA2008U_120 A SCHEDULING REINFORCEMENT LEARNING ALGORITHM

Improving Conceptual Understanding of Physics with Technology

B.S/M.A in Mathematics

The Indices Investigations Teacher s Notes

Generating Test Cases From Use Cases

DOCTOR OF PHILOSOPHY HANDBOOK

Number Line Moves Dash -- 1st Grade. Michelle Eckstein

FOR TEACHERS ONLY. The University of the State of New York REGENTS HIGH SCHOOL EXAMINATION PHYSICAL SETTING/PHYSICS

An Effective Framework for Fast Expert Mining in Collaboration Networks: A Group-Oriented and Cost-Based Method

SEMAFOR: Frame Argument Resolution with Log-Linear Models

Introduction to Causal Inference. Problem Set 1. Required Problems

Continual Curiosity-Driven Skill Acquisition from High-Dimensional Video Inputs for Humanoid Robots

Copyright 2017 DataWORKS Educational Research. All rights reserved.

Assessing Children s Writing Connect with the Classroom Observation and Assessment

The Strong Minimalist Thesis and Bounded Optimality

Machine Learning from Garden Path Sentences: The Application of Computational Linguistics

Enduring Understandings: Students will understand that

Community Power Simulation

Evolutive Neural Net Fuzzy Filtering: Basic Description

Undergraduate Program Guide. Bachelor of Science. Computer Science DEPARTMENT OF COMPUTER SCIENCE and ENGINEERING

Rover Races Grades: 3-5 Prep Time: ~45 Minutes Lesson Time: ~105 minutes

Learning to Schedule Straight-Line Code

Major Milestones, Team Activities, and Individual Deliverables

FUZZY EXPERT. Dr. Kasim M. Al-Aubidy. Philadelphia University. Computer Eng. Dept February 2002 University of Damascus-Syria

TA Script of Student Test Directions

Given a directed graph G =(N A), where N is a set of m nodes and A. destination node, implying a direction for ow to follow. Arcs have limitations

Facing our Fears: Reading and Writing about Characters in Literary Text

Seminar - Organic Computing

Honors Mathematics. Introduction and Definition of Honors Mathematics

Grade 4: Module 2A: Unit 2: Lesson 4 Word Choice: Using Academic Vocabulary to Apply for a Colonial Trade Job

Navigating the PhD Options in CMS

GACE Computer Science Assessment Test at a Glance

Prediction of Maximal Projection for Semantic Role Labeling

Spring 2015 IET4451 Systems Simulation Course Syllabus for Traditional, Hybrid, and Online Classes

Predicting Students Performance with SimStudent: Learning Cognitive Skills from Observation

A General Class of Noncontext Free Grammars Generating Context Free Languages

Module 12. Machine Learning. Version 2 CSE IIT, Kharagpur

Creating a Test in Eduphoria! Aware

Cal s Dinner Card Deals

Business 712 Managerial Negotiations Fall 2011 Course Outline. Human Resources and Management Area DeGroote School of Business McMaster University

Some Principles of Automated Natural Language Information Extraction

"f TOPIC =T COMP COMP... OBJ

Multimedia Application Effective Support of Education

Innovative Methods for Teaching Engineering Courses

What s in a Step? Toward General, Abstract Representations of Tutoring System Log Data

CUNY ASSESSMENT TESTS Webinar for International Students

Automatic Discretization of Actions and States in Monte-Carlo Tree Search

High-level Reinforcement Learning in Strategy Games

Section 6 DISCIPLINE PROCEDURES

Regret-based Reward Elicitation for Markov Decision Processes

Transfer Learning Action Models by Measuring the Similarity of Different Domains

Objectives. Chapter 2: The Representation of Knowledge. Expert Systems: Principles and Programming, Fourth Edition

Transcription:

CSE 473 Autumn 2012: Sample Midterm Exam Solutions (closed book, closed notes except for 1-page summary) Total: 100 points, 5 questions. Time: 50 minutes Instructions: 1. Write your name and student ID on the first sheet (once you start, write your last name on all sheets). Write your answers in the space provided. If you need more space or scratch paper, you can get additional sheets from the instructor. Make sure you write down the question number and your name on any additional sheets. 2. Read all questions carefully before answering them. Feel free to come to the front to ask for clarifications. 3. Hint 1: You may answer the questions in any order, so if you find that you're having trouble with one of them, move on to another one. 4. Hint 2: If you don t know the answer to a question, don t omit it - do the best you can! You may still get partial credit for whatever you wrote down. Good luck! Do not start until you are given the go-ahead signal 0

1) (20 points) Agents and Environments a) Give a PEAS description of the task environment for the following agents: i) A part-picking robot ii) An interactive English tutor b) For each of the agents above, characterize the environment according to whether it is fully or partially observable, deterministic or stochastic, episodic or sequential, static or dynamic, discrete or continuous, and single or multiagent. [For a solution, see Figures 2.5 and 2.6 in the text] 1

2) (20 points) Uninformed Search Consider a state space where the start state is 1 and the successor function for state i returns two states: 2i, 2i+1. a) Draw the state space graph for states 1 to 15. b) Suppose the goal state is 11. List the order in which nodes will be visited for: (i) breadth-first search, (ii) depth-limited search with depth limit 3, and (iii) iterative deepening search. [Solution: a) Follow the format of the state space graph in Lecture 3, starting at 1 with arrows to 2 and 3, etc. b) (i) 1, 2, 3,, 10, 11 (ii) 1, 2, 4, 8, 9, 5, 10, 11 (iii) 1, 1, 2, 3, 1, 2, 4, 5, 3, 6, 7, 1, 2, 4, 8, 9, 5, 10, 11.] 2

3) (20 points) Informed Search Consider the route-finding problem of computing the cheapest path from Arad to Bucharest in the map depicted above. For the straight-line distance heuristic, draw the search tree after expansion of each node until the termination of the algorithm for: a) Greedy best-first search (label all nodes with their h values). What is the solution found by the algorithm? b) A* search (label all nodes with their f values). What is the solution found by the algorithm? [For the solution, see lecture slides on Informed Search or Figures 3.23/3.24 in text] 3

4) (15 points) Adversarial Search Consider the 4-ply game tree above for a two-person game. a) Fill in the minimax values of all nodes marked 0. b) Label the path in the tree that results if max and min each make optimal moves. c) Draw the tree that results if alpha-beta pruning is used. [For the solution, see lecture slides on Alpha-Beta Search] 4

5) (25 points) Logic Any American who sells weapons to hostile nations is a criminal. The nation Nono is an enemy of America. Any enemy of America is considered hostile. Nono has some missiles. Missiles are weapons. All of Nono s missiles were sold to it by Colonel West. West is American. a) Convert the statements above to sentences in first-order logic. There should be seven sentences. Use the following vocabulary: American(x), Weapon(x), Hostile(x), Criminal(x), Missile(x): predicates satisfied by members of the corresponding categories Sells(x,y,z): x sells y to nation z Owns(x,y): nation x owns y Enemy(x,y): nation x is an enemy of nation y West: constant denoting Colonel West Nono, America: constants denoting corresponding nations b) First, make sure your sentences in (a) are definite clauses (if not, make them so). Then, using forward chaining, infer that West is a criminal: draw the proof tree with initial facts at the bottom, facts inferred on the 1 st iteration at the next level, and so on. c) Convert the sentences in (a) to CNF. d) Using the CNF sentences in (c), construct a resolution proof that West is a criminal. [For the solution, see Section 9.3.1, Figure 9.4, Section 9.5.3, and Figure 9.11 in the text] 5