CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 7: Methods for Index Selection The indexes of a case allow us to retrieve it when we need.

Slides:



Advertisements
Similar presentations
For a healthy heart…watch the salt Looking to enhance your health? Then watch the salt…. Eating a diet high in salt can increase your blood pressure. Higher.
Advertisements

Lecture 5: Reuse, Adaptation and Retention
Casserole A casserole is a tasty blend of cooked ingredients that are heated together to develop flavor.
BOILED OMELETTESCRAMBLED FRIED POACHED Soft boiled eggs; the yolk, and sometimes even the white, remains "runny". Hard-boiled eggs are produced by boiling.
Machine Learning Instance Based Learning & Case Based Reasoning Exercise Solutions.
Easy-to-Use Online Health Management Services A secure connection between you and your clients.
Stir-Fries and Casseroles. Objective Compare stir-fries with casseroles Describe how to prepare ingredients for a stir-fry dish Demonstrate how to cook.
Banquet and Buffet Food Production
Basic Culinary Skills Unit 7. Know about purchasing food and the associated documents Purchasing is the technical term used in catering for buying. Hospitality.
Machine Learning and Data Mining Clustering
CHEF (Hammond, 1987) CHEF is a case-based planner, which can output new recipes given particular ingredients and tastes. Recipes are viewed as plans. They.
Case-Based Reasoning Ramon López de Mántaras Badia IIIA - CSIC
Chapter 2 Simple Linear Regression – What is it, Why do we do it?
Front Cover Design a front cover for your assignment make sure you have; Pasta Sauce Design Task, a nice picture and your name.
Eating value for your dollar.  Eating a healthy diet doesn’t have to cost a fortune.  In fact you can eat delicious healthy food and save money!  By.
Sepandar Sepehr McMaster University November 2008
Algebra 1 R. Jenkins, M.S., M.A..
American Food. Table of contents  Main points of american cooking Main points of american cooking  Fast Food Fast Food  Apple pie Apple pie  Buffalo.
Healthy Eating on a Budget Eating value for your dollar.
School of Computer Science and Technology, Tianjin University
CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 3: Reasoning Using Cases In this chapter, we look at how cases are used to reason We’ve already.
CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 2: Case Studies of Several Case-Based Reasoners This chapter describes some classic CBR systems.
Taking Food Order Meeting 4 Subject: V Restaurant English Year:
Lecture2: Database Environment Prepared by L. Nouf Almujally & Aisha AlArfaj 1 Ref. Chapter2 College of Computer and Information Sciences - Information.
Participate in a Team to Achieve Organizational Goal
Case study of Several Case Based Reasoners Sandesh.
Today we will compute a given percent of a whole number. compute – figure out.
Applied Software Project Management
CMPF144 FUNDAMENTALS OF COMPUTING THEORY Module 10: Applied Mathematical - Logical Knowledge Riddle/Puzzle.
The Scientific Method Chapter 5.
1. What is a recipe? A blueprint or a pattern to follow in preparing foods.
GroceryList++ (MEAL BASED SHOPPING LIST). Team Members Staci Menz  Major: Computer Science  Outside Interests: Video games, music, reading Brian Chan.
CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 11: Adaptation Methods and Strategies Adaptation is the process of modifying a close, but.
CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 8: Organizational Structures and Retrieval Algorithms This chapter deals with how to find.
Chapter 11 Planning Meals
1 Knowledge Acquisition and Learning by Experience – The Role of Case-Specific Knowledge Knowledge modeling and acquisition Learning by experience Framework.
CHALLENGES. Solve the following challenges: 1. You have prepared your recipe for 234 people but in the last minute the chef says that you need to distribute.
SQL Based Knowledge Representation And Knowledge Editor UMAIR ABDULLAH AFTAB AHMED MOHAMMAD JAMIL SAWAR (Presented by Lei Jiang)
Menus.
1) Personal hygiene preparation. 2) Collect recipe & timeplan. 3) Cook!
CS 1120: Computer Science II Software Life Cycle Slides courtesy of: Prof. Ajay Gupta and Prof. James Yang (format and other minor modifications by by.
Appetizers, Soups and Salad 1. Norwegian Smoked Salmon 2. Baked Lobster Bisque 3. Onion Soup 4. Garden Green Salad 5. Chef’s Salad  Your choice of.
CS223: Software Engineering Lecture 3: Software Develoment Processes.
CS CS1321: Introduction to Programming Georgia Institute of Technology College of Computing Lecture 6 Sept 6th, 2001 Fall Semester.
1 Getting Ready to Cook Chapter
1 Planning Meals at Home Chapter Taste Buds Detect only 4 basic sensations Salty Sweet Bitter Sour.
MANAGEMENT INFORMATION SYSTEMS (MIS) AND OTHER INFORMATION SYSTEMS.
Meal Planning Just FACS. What’s for breakfast, lunch and dinner ?
Note Taking Strategies Top 5 Tips for Effective Note Taking.
1 Casseroles, Stir-Fries, etc Chapter Casserole = mixture of cooked foods heated in a baking or casserole dish.
Inductive and Deductive Reasoning  The pre-requisites for this chapter have not been seen since grade 7 (factoring, line constructions,..);
WHAT ARE PLANS FOR? Philip E. Agre David Chapman October 1989 CS 790X ROBOTICS Presentation by Tamer Uz.
A set of directions for preparing a food product. Recipe:
#1 Make sense of problems and persevere in solving them How would you describe the problem in your own words? How would you describe what you are trying.
UNIT III. A managerial problem can be described as the gap between a given current state of affairs and a future desired state. Problem solving may then.
Using Recipes Essential Question: How does an understanding of the components of a recipe help you produce successful results? Food for Today Chapter 24.
CHAPTER Images shutterstock.com 10 Planning Meals.
Nutritional Balance for You and the Person You Are Caring For Lauren Gogol, MScFN, RD April 19 th, 2016 Care for the Caregiver Information Series.
THE ITALIAN CUISINE Заикина Ольга Петровна Учитель английского языка
Cost Control Objectives: What is the importance of cost control?
Vegetable Main Dish Assignment
Chapter 9: Matching and Ranking Cases
Optional Unit 5 Preparing Food for a Specified Function
UNIT 3 CHAPTER 1 LESSON 4 Using Simple Commands.
CS 1120: Computer Science II Software Life Cycle
Physician Diagnosis and Prescription
FRONT No Solutions Infinite Solutions 1 solution (x, y)
CS 1120: Computer Science II Software Life Cycle
Chapter 2 Section 3.
Optional Unit 5 Preparing Food for a Specified Function
Presentation transcript:

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling1 Chapter 7: Methods for Index Selection The indexes of a case allow us to retrieve it when we need it again A useful index has four properties It should be predictive It should make useful predictions It should be easy to recognize It should be generally applicable Indexes are typically specialized for the problem domain Attempts to automate or generalize index selection only work in the simplest of problem domains

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling2 Index Selection: CHEF Example An index may be stored as part of a case or may be external to the case Consider CHEF A case contains a problem and a solution The problem is the goals you want to achieve, like making a stir-fry dish that contains beef and broccoli The solution includes the ingredients for the dish and the steps you need to take to prepare the dish Some indexes, like contains-beef and contains-broccoli, are parts of the case Other predictive indexes are not part of the case itself Contains-meat is more general than contains-beef. It allows us to predict that a vegetarian won’t like the dish Avoids-soggy-broccoli-problem allows us to make predictions about other recipes in which vegetables could get soggy

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling3 Steps in Index Selection To choose good indexes for a case: Determine what tasks the case is useful for Determine when the case is likely to be useful Choose features of the case that predict the case will be useful for those tasks under those circumstances Make those features more general, if possible Describe those features using vocabulary your computer program can recognize and process

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling4 Index Selection: JULIA Example Problem: 20 people were coming to dinner It was summer Tomatoes were in season The meal should be vegetarian One guest was allergic to milk Solution: Serve tomato tart (a cheese and tomato pie) Adapt one tart by using tofu cheese substitute instead of cheese

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling5 Index Selection: JULIA Example (continued) This case is useful for: Finding a vegetarian main dish with tomatoes Adapting a dish with cheese when you have to eliminate milk products The tasks this case can help with are: Choosing a main dish Adapting a dish to eliminate cheese The circumstances under which this case will be useful are: When you need to choose a vegetarian main dish in summer When you need to eliminate cheese from a main dish Note that you need to think about the case from all angles, even if it seems like you’re just saying the same thing in different ways

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling6 Indexes in JULIA

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling7 Indexes in JULIA (continued)

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling8 Checklist-Based Indexes Some systems are able to index all cases by the same features The analysis is done up front in figuring out which features are predictive When it works, this facilitates later reasoning, because the system can always look in the same slots to find predictive features CASEY could use a checklist approach, because the patient’s symptoms were always what was needed to predict that one patient’s diagnosis would be useful for another patient CHEF could get some, but not all, of its indexes this way Cases were always indexed on major ingredients and preparation method Other indexes were case specific

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling9 Difference-Based Indexing You can sometimes reduce the number of indexes by considering how well different features differentiate cases from each other For example, if you know a patient’s temperature is predictive, but most patients have a normal temperature, you could use temperature as an index only when a patient has a fever Depending on your retrieval approach, having fewer indexes may improve efficiency Note: In a diagnostic system, differentiating features and predictive features are exactly the same thing. In our system to prescribe neuroleptics for Alzheimer’s Disease patients, most of the effort required was spent determining which features distinguished patients who took neuroleptics from those who did not Patient charts contained hundreds of features. A physician recommended the top 100 features, and we narrowed this down to ten predictive indexes

CS 682, AI:Case-Based Reasoning, Prof. Cindy Marling10 Explanation-Based Indexing If a system is able to generate explanations of why a solution did or did not work, these explanations make useful indexes This is what CHEF used to get indexes like avoids-soggy-vegetable- problem This isn’t possible in domains that are not well understood For example, Lockheed engineers never knew why one way of loading the autoclave worked and another didn’t When a system (or a person) can create good explanations, then this is a good way to index cases