Date post: | 09-Jun-2015 |
Category: |
Engineering |
Upload: | amironline |
View: | 235 times |
Download: | 0 times |
A M I R B A D A M C H I
CASE-BASED REASONING CASE STUDY: HOUSING PRICE
Amirkabir University of Technology
Computer Engineering & Information
Technology Faculty
CONTENTS
• CBR
• Definition
• Assumptions
• Cycle
• Advantages, disadvantages
• Housing Price
• Introduction
• Method
• Estimation Function
• Similarity Function
• Results
DEFINITION
• Case-based reasoning is […] reasoning by remembering - Leake, 1996
• A case-based reasoner solves new problems by adapting solutions that were used to solve old problems - Riesbeck & Schank, 1989
• Case-based reasoning is a recent approach to problem solving and learning […] - Aamodt & Plaza, 1994
CBR ASSUMPTIONS
• The main assumption is that:• Similar problems have similar solutions:
• e.g., an aspirin can be taken for any mild pain
• Two other assumptions:• The world is a regular place: what holds true today will
probably hold true tomorrow • (e.g., if you have a headache, you take aspirin, because it has
always helped)
• Situations repeat: if they do not, there is no point in remembering them • (e.g., it helps to remember how you found a parking space near
that restaurant)
CBR CYCLE
• Retrieve:• Determine most similar case(s).
• Reuse:• Solve the new problem re-using information and
knowledge in the retrieved case(s).
• Revise:• Evaluate the applicability of the proposed
solution in the real-world.
• Retain:• Update case base with new learned case for
future problem solving
CBR CYCLE
ADVANTAGES
• solutions are quickly proposed• derivation from scratch is avoided
• domains do not need to be completely understood
• cases useful for open-ended/ill-defined concepts
• highlights important features
DISADVANTAGES
• old cases may be poor
• library may be biased
• most appropriate cases may not be retrieved
• retrieval/adaptation knowledge still needed
HOUSING PRICE
• Mary wishes to sell her apartment in the city.
• She might start with the price she paid for her
apartment and add an annual appreciation that
seems reasonable to her.
• She might try to predict market trends and figure
out how much the apartment should be worth.
HOUSING PRICE
• General rules
• In this area, the price per squared meter is $3,000..
• Case-based
• The apartment next door, practically identical to mine, was
just sold for $300,000..
METHOD
METHOD(CONT)
ESTIMATION FUNCTION
• Use parametric approach
• Advantages:
• Simplify to analyse
• Comparasion of two models
• Hyphotehes Test
SIMILARITY FUNCTION
• Weighted euclidean distance
• Why weighted euclidean distance instead standard
euclidean distance
• Variables with differenet scales
• Variables with differenet influent
• Allow a wide range of distance functions, weighing the
relative importance of variables
SIMILARITY FUNCTION
• Translate the distance function to a similarity
function
• decreasing in the distance
• The distance goes up from 0 to
• The similarity function goes down from 1 (maximal
similarity) to 0.
RESULTS
• Goodness of fit measures for regression and
similarity, for the two databases.
LIKE :Value of the log-likelihood function (in-sample, 75% of the data
points)
SSPE: Sum of Squared Prediction Errors (out of sample, remaining 25%
of the data points)
AIC: Akaike Information Criterion (computed over the whole sample)
SC : Schwarz Criterion (computed over the whole sample)
REFERENCES
• Ian Watson, “An Introduction to Case-Based
Reasoning”, 1995.
• Gayer, Gilboa, Lieberman,"Rule-Based and Case-
Based Reasoning in Housing Prices", 2007.
• Billot, A., I. Gilboa, D. Samet, and D. Schmeidler,
"Probabilities as Similarity-Weighted Frequencies",
2005.
Any question?
THAT’S ALL FOLKS!