61
1 2, G.Tecuci, Learning Agents Laboratory Learning Agents Laboratory Computer Science Department George Mason University Prof. Gheorghe Tecuci Learning by analogy

09 analogical-learning

Embed Size (px)

Citation preview

Page 1: 09 analogical-learning

1 2002, G.Tecuci, Learning Agents Laboratory

Learning Agents LaboratoryComputer Science Department

George Mason University

Prof. Gheorghe Tecuci

Learning by analogy

Page 2: 09 analogical-learning

2 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 3: 09 analogical-learning

3 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: definition

Learning by analogy means acquiring new knowledge about an input entity by transferring it from a known similar entity.

Which is the central intuition supporting the learning by analogy paradigm?

Qa=3 Qb=9

Qc=?

Simple Hydraulics Problem

I1 I2

I3=I1+I2

Kirchoff's First Law

One may infer, by analogy, that hydraulics laws are similar to Kirchoff's laws, and Ohm's law.

Page 4: 09 analogical-learning

4 2002, G.Tecuci, Learning Agents Laboratory

Discussion

Central intuition supporting learning by analogy:If two entities are similar in some respects then they could be similar in other respects as well.

Examples of analogies:

Pressure Drop is like Voltage Drop

A variable in a programming language is like a box.

Provide other examples of analogies.

Page 5: 09 analogical-learning

5 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: illustration

sun

planet

yellow

mass

mass

temperature

greater

color

revolves-around

attractsTsun

Tplanet

Msun

Mplanet

causes

temperature

greater

mass

mass

revolves-around

attractsMnucleus

causes ?greater

nucleus

electron

Melectron

sun

planet

yellow

mass

mass

temperature

greater

color

revolves-around

attractsTsun

Tplanet

Msun

Mplanet

causes

temperature

greater

mass

mass

revolves-around

attractsMnucleus

causes ?greater

nucleus

electron

Melectron

Illustration: The hydrogen atom is like our solar system.

The Sun has a greater mass than the Earth and attracts it, causing the Earth to revolve around the Sun. The nucleus also has a greater mass then the electron and attracts it.Therefore it is plausible that the electron also revolves around the nucleus.

Page 6: 09 analogical-learning

6 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: the learning problem

Given:

• A partially known target entity T and a goal concerning it.

• Background knowledge containing known entities.

Find:

• New knowledge about T obtained from a source entity S belonging to the background knowledge.

Partially understood structure of the hydrogen atom under study.

Knowledge from different domains, including astronomy, geography, etc.

In a hydrogen atom the electron revolves around the nucleus, in a similar way in which a planet revolves around the sun.

Page 7: 09 analogical-learning

7 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: the learning method• ACCESS: find a known entity that is analogous with the input entity.

• MATCHING: match the two entities and hypothesize knowledge.

• EVALUATION: test the hypotheses.

• LEARNING: store or generalize the new knowledge.Store that, in a hydrogen atom, the electron revolves around the nucleus. By generalization from the solar system and the hydrogen atom, learn the abstract concept that a central force can cause revolution.

In the Rutherford’s analogy the access is no longer necessary because the source entity is already given (the solar system).

One may map the nucleus to the sun and the electron to the planet, allowing one to infer that the electron revolves around the nucleus because the nucleus attracts the electron and the mass of the nucleus is greater than the mass of the electron.

A specially designed experiment shows that indeed the electron revolves around the nucleus.

Page 8: 09 analogical-learning

8 2002, G.Tecuci, Learning Agents Laboratory

Discussion

How does analogy help?

Why not just study the structure of the hydrogen atom to discover that new knowledge?We anyway need to perform an experiment to test that the electron revolves around the hydrogen atom.

Page 9: 09 analogical-learning

9 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 10: 09 analogical-learning

10 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: Design issues

• ACCESS: find a known entity that is analogous with the input entity.

• MATCHING: match the two entities and hypothesize knowledge.

• EVALUATION: test the hypotheses.

• LEARNING: store or generalize the new knowledge.How to learn?

Given a target, how to identify a few potential sources in a very large storage?

Given a potential source, how to identify the knowledge to hypothesize?

Why and how to test the hypothesized knowledge?

Page 11: 09 analogical-learning

11 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: Formalization

Given: - a target entity T; - a universe of potential sources U; - an access function f1 with a threshold value - a matching function f2 with a threshold value .

Find: - new knowledge about T (using analogical learning).

Page 12: 09 analogical-learning

12 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: Access

Find potential sources for T in U :

f1(Sk, T) >

This should result in S1, … , Sn

Page 13: 09 analogical-learning

13 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: Matching

Find the best match between one of S1, …, Sn and T.

Let: Sk = A & B & C, T = A' & Dwhere f2(Sk, T) > 2 gives the best match.

A and A' are the parts of Sk and T that make them analogous:f2(Sk, T) = f2(A, A')

B, C and D are the other parts of Sk and T.

As a side effect of partially matching Sk with T (or totally matching A with A'), one obtains a correspondence (substitution) list = ( o1 o1', ... , on on') where oi is an element of A and oi' is the corresponding element from A'.By applying the substitution to Sk one obtains:

(Sk) = (A) & (B) & (C) = A' & (B) & (C) = A' & B' & C'.

By analogy with Sk one concludes that T might also have the features B' & C'.

Page 14: 09 analogical-learning

14 2002, G.Tecuci, Learning Agents Laboratory

Learning by analogy: Evaluation and learning

By analogy with Sk one concludes that T might also have the features B' & C'.

However, the evaluation phase shows that T has the features B' but it does not have the features C'.

Therefore:- B represents the part of Sk that is transferred to T because of the similarity between A and A‘;- C is the part of Sk that is not transferred to T;- D represents the features that are specific to T.

Page 15: 09 analogical-learning

15 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: Rutherford’s analogy

In this case, the fact that S and T are analogous is already known. Therefore, the access part is solved and the only purpose of the matching function remains that of identifying the correct correspondence between the elements of the solar system and those of the hydrogen atom.

This is an example of a special (simpler form of analogy):

“A T is like an S.”

This is useful mostly in teaching based on analogy.

"The hydrogen atom is like our solar system".

Page 16: 09 analogical-learning

16 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: potential matchings

Which are the possible matchings between the elements of S and the elements of T?

sun

planet

yellow

mass

mass

temperature

greater

color

revolves-around

attractsTsun

Tplanet

Msun

Mplanet

causes

temperature

greater

mass

mass

attractsMnucleus

greater

nucleus

electron

Melectron

sun

planet

yellow

mass

mass

temperature

greater

color

revolves-around

attractsTsun

Tplanet

Msun

Mplanet

causes

temperature

greater

mass

mass

attractsMnucleus

greater

nucleus

electron

Melectron

Page 17: 09 analogical-learning

17 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: potential matchingsThere are several possible matchings between the elements of S and the elements of T and one has to select the best one:

Matching1:sun nucleus, planet electron, Msun Mnucleus, Mplanet Melectron,

which is supported by the following correspondencesmass(sun, Msun) mass(nucleus, Mnucleus)mass(planet , Mplanet ) mass(electron, Melectron)greater(Msun, Mplanet) greater(Mnucleus, Melectron),attracts(sun, planet) attracts(nucleus, electron)

Matching2:sun nucleus, planet electron, Tsun Mnucleus, Tplanet Melectron,

that is supported by the following correspondencesgreater(Tsun, Tplanet) greater(Mnucleus, Melectron),attracts(sun, planet) attracts(nucleus, electron)

Matching3:sun electron, planet nucleus, Msun Melectron, Mplanet Mnucleus

Page 18: 09 analogical-learning

18 2002, G.Tecuci, Learning Agents Laboratory

Similarity estimation issues and sample solutions

4. How to define the similarity threshold ?

The similarity of two entities is the sum of the similarity of their elements.Other solutions?

Exhaustive search.Other solutions?

Two elements are similar if they represent the same concept or are subconcepts of the same concepts. In such a case their similarity may be considered 1 (on a 0-1 scale).Other solutions?

Similarity threshold defined by the designer (a hard critical issue).Other solutions?

1. How to search the space of all possible matchings ?

2. How to measure the similarity of two elements ?

3. How to combine the estimated similarities of the parts in order to obtain the similarity between S and T ?

Page 19: 09 analogical-learning

19 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: Matching result

The best matching is Matching1 (because it leads to the highest number of common features of the solar system and the hydrogen atom) that gives the following substitution: = (sun nucleus, planet electron, Msun Mnucleus, Mplanet Melectron)

The features in light color are those that could be transferred to the hydrogen atom as a result of the analogy with the solar system:

• color(nucleus, yellow)• temperature(nucleus, Tn)• temperature(electron, Te)• greater(Tn, Te)• revolves-around(nucleus, electron)• causes( (attracts(nucleus,electron),

greater(Mnucleus, Melectron)),revolves-around(nucleus, electron))

yellow

mass

mass

temperature

greater

color

revolves-around

attractsTsun

Tplanet

causes

temperature

greater

Mnucleus

nucleus

electron

Melectron

yellow

-

Tnucleus

Telectron

temperatureelectron

By applying the substitution to the solar system, one obtains the following structure:

Page 20: 09 analogical-learning

20 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: Evaluation

The evaluating phase shows that

The hydrogen atom has the features:• revolves-around(nucleus, electron)• causes((attracts(nucleus,electron), greater(Mnucleus, Melectron)),

revolves-around(nucleus, electron))

The hydrogen atom does not have the features:• color(nucleus, yellow)• temperature(nucleus, Tn)• temperature(electron, En)• greater(Tn, En)

Which is, in your opinion, the most critical issue in analogical learning?

Page 21: 09 analogical-learning

21 2002, G.Tecuci, Learning Agents Laboratory

Discussion

What kind of features may be transferred from the source to the target so as to make sound analogical inferences ?

Which is the most critical issue in analogical learning?

Page 22: 09 analogical-learning

22 2002, G.Tecuci, Learning Agents Laboratory

attracts(sun,planet) mass(sun,Msun) mass(planet,Mplanet) greater(Msun,Mplanet)

(sun <- planet, planet <- electron, Msun <- Mnucleus, Mplanet <- Melectron)

=

attracts(nucleus,electron) mass(nucleus,Mnucleus) mass(electron,Melectron) greater(Mnucleus,Melectron)

revolves-arround(planet,sun) revolves-arround(electron,nucleus)

CAUSE CAUSE ?

Case study discussion: transfer of causal relation

Page 23: 09 analogical-learning

23 2002, G.Tecuci, Learning Agents Laboratory

Case study discussion: Learning

Store the new acquired knowledge about the hydrogen atom:

• revolves-around(nucleus, electron)• causes(attracts(nucleus,electron), greater(Mnucleus, Melectron)),

revolves-around(nucleus, electron))

By generalization from the solar system and the hydrogen atom one may learn the abstract concept that a central force can cause revolution:

• causes(attracts(x, y), greater(Mx, My)), revolves-around(x, y))

Question:

When to store the acquired knowledge and when to generalize it?

Page 24: 09 analogical-learning

24 2002, G.Tecuci, Learning Agents Laboratory

Analogy in Disciple

similar exampleexplains?

similarIdentify and test a strategic COG candidate for a force The force is Germany_1943

I need to

Therefore I need to

Identify and test a strategic COG candidate corresponding to a member of a force The force is European_Axis_1943

initial example

explanation

explains

Identify and test a strategic COG candidate for a force The force is US_1943

I need to

Therefore I need to

Identify and test a strategic COG candidate corresponding to a member of a force The force is Allied_Forces_1943

US_1943has_as_memberAllied_Forces_1943

similar explanationless general than

Analogycriterion

less general than

?O2has_as_member?O1

forcemulti_member_force

instance_ofinstance_of

Germany_1943has_as_memberEuropean_Axis_1943

similar

Page 25: 09 analogical-learning

25 2002, G.Tecuci, Learning Agents Laboratory

A A'

B B'

similar

similar

causes causes ?The basic scheme of analogy

Causal networks of relationsAn important result of the learning by analogy research is that the analogy involves mapping some underlying causal network of relations between analogous situations.

By causal network of relations it is generally meant a set of relations related by special higher order relations such as 'physical-cause(ri, rj)', 'logically-implies(ri, rj)', 'enables(ri, rj)', 'justifies(ri, rj)', determines(ri, rj) etc.

The idea is that similar causes are expected to have similar effects:

Page 26: 09 analogical-learning

26 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 27: 09 analogical-learning

27 2002, G.Tecuci, Learning Agents Laboratory

Gentner’s structure mapping theory

The main claim of this theory is that relations between objects, rather than attributes of objects, are mapped from source to target. Moreover, a relation that belongs to a mappable system of mutually interconnecting relationships is more likely to be imported into the target than is an isolated relation (the systematicity principle).

See:Gentner D., The mechanisms of analogical reasoning, in J.W.Shavlik, T.G.Dietterich (eds), Readings in Machine Learning, Morgan Kaufmann, 1990.

Page 28: 09 analogical-learning

28 2002, G.Tecuci, Learning Agents Laboratory

Gentner’s structure mapping theory (cont.)

Analogy maps the objects of the source onto the objects of the target: s1 t1, ... , sn tn

These object correspondences are used to generate the candidate set of inferences in the target domain.

Predicates from the source are carried across to the target, using the node substitutions dictated by the object correspondences, according to the following rules:

1. Discard attributes of objects A(si) -/-> A(ti)For instance, the yellow color of the sun is not transferred to the hydrogen nucleus.

2. Try to preserve relations between objects R(si, sj) -?-> R(ti, tj)That is, some relations are transferred to the target, while others are not.

3. The systematicity principle: the relations that are most likely to be transferred are those belonging to systems of interconnected relations

R'(R1(si,sj), R2(sk,sl)) R'(R1(ti,tj), R2(tk,tl))

Page 29: 09 analogical-learning

29 2002, G.Tecuci, Learning Agents Laboratory

Literal similarity, analogy, and abstraction

Gentner's theory distinguishes between literal similarity, analogy, and abstraction.

One says that a target T is literally similar with a source S if and only if a large number of predicates is mapped from source to target, relative to the number of nonmapped predicates and, also, the mapped predicates include both attributes of objects and relations between objects.

For instance, 'kool-aid' is literally similar with 'water' since it has most of the features of 'water' (both attributes of objects and relations between objects).

Give other examples of literally similar entities.

Page 30: 09 analogical-learning

30 2002, G.Tecuci, Learning Agents Laboratory

Literal similarity, analogy, and abstraction

One says that a target T is analogous with a source S if and only if relations between objects, but few or no attributes of objects, can be mapped from source to target.

For instance, 'heat' is analogous to 'water'.

Give other examples of abstractions.

One says that a source S is an abstraction of a target T if and only if the source is an abstract relational structure and each predicate (a relation between objects or an attribute of an object) from the abstract source is mapped into a less abstract predicate of the target; there are no nonmapped predicates.

For instance, 'through-variable' is an abstraction of 'heat', where by 'through-variable' we mean something that flows across a difference in potential.

Page 31: 09 analogical-learning

31 2002, G.Tecuci, Learning Agents Laboratory

Similarity, analogy, and abstraction: discussion

Given that two entities overlap in relations, they are more literally similar to the extent that their object attributes also overlap. Therefore, literal similarity might be seen as a particular case of analogy.

Abstraction may also be seen as a special case of analogy in which all the predicates of the source entity are mapped into the target entity.

What could we conclude from these observations?

Page 32: 09 analogical-learning

32 2002, G.Tecuci, Learning Agents Laboratory

Similarity, analogy, and abstraction: discussion

The contrast between literal similarity, analogy, and abstraction is a continuum.

What could we conclude from these observations?

Overlap in relations is necessary for any perception of similarity, analogy or abstraction.

Page 33: 09 analogical-learning

33 2002, G.Tecuci, Learning Agents Laboratory

Gentner’s theory: implementation and discussion

An implementation of the Structure-Mapping theory is the Structure-Mapping Engine (Falkenhainer, Forbus & Gentner, 1989: The Structure-mapping Engine. Algorithms and Examples, Artificial Intelligence, 41:1-63. Also in Readings in Knowledge Acquisition and Learning).

Given the descriptions of a source and a target, the Structure-Mapping Engine constructs all syntactically consistent analogical mappings between them. Each mapping consists of pairwise matches between predicates and objects in the source and target, plus a list of predicates which exist in the source but not the target. This list of predicates is the set of candidate inferences sanctioned by the analogy. The Structure-Mapping Engine evaluates syntactically each possible analogy to find the best one.

Page 34: 09 analogical-learning

34 2002, G.Tecuci, Learning Agents Laboratory

Gentner’s theory: implementation and discussion

The Structure-Mapping Engine needs to be given the descriptions of a source and a target. This requires the ACCESS problem to be solved first:

How do we find potential sources for a target?

MAC/FAC (Forbus, Gentner, Law, 1995: “MAC/FAC: A model of similarity-based retrieval,” Cognitive Science, 19(2):141-205) is a system that addresses the access problem.

The MAC stage uses a simple, nonstructural matcher to filter our a few promising candidates from a large memory of structured descriptions.The FAC stage evaluates each candidate using SME to provide a structural match.

MAC/FAC was scaled-up in the DARPA’s HPKB and RKF programs.

What is, however, a problem with Gentner’s theory?

Page 35: 09 analogical-learning

35 2002, G.Tecuci, Learning Agents Laboratory

Gentner’s theory: discussion

Gentner’s interpretation rules depend only on the syntactic properties of the knowledge representation, and not on the specific content of the domain.

What is a problem with Gentner’s theory?

Why is this a problem?Consider these equivalent representations:

Book1-on-TableOn(Book1, Table)

causes((attracts(nucleus,electron), greater(Mnucleus, Melectron)),revolves-around(nucleus, electron))

Could you think of a different representation where the following expression is no longer a second order relation?

Page 36: 09 analogical-learning

36 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 37: 09 analogical-learning

37 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Definition

Instead of giving a general criterion for the validity of analogical knowledge transfer (high order relations or causal network of relations), Russel and Davis propose to specify explicitly what knowledge can be transferred. The rules for specifying this are called "determination rules".

P(x, y) >- Q(x, z) (P plausibly determines Q) meaning

S, T { If y [P(S, y) & P(T, y)] then it is probably true that z [Q(S, z) & Q(T, z)] }

where P and Q are first order logical expressions.

Page 38: 09 analogical-learning

38 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Definition (cont.)

A determination rule is an expression of the following form:U(x1,...,xn,y1,...,ym) >- V(x1,...,xn,z1,...,zp)

One says that U determines V. That is, whenever the arguments of U have certain values, the arguments of V are very likely to have corresponding values.

Example: Rainfall(x, y) >- Water-in-soil(x, z)Rainfall(Philippine, heavy), Water-in-soil(Philippine, high)

Page 39: 09 analogical-learning

39 2002, G.Tecuci, Learning Agents Laboratory

Given: Rainfall(x, y) >- Water-in-soil(x, z)Rainfall(Philippine, heavy), Water-in-soil(Philippine, high)Rainfall(Vietnam, heavy)

Conclude: Water-in-soil(Vietnam, high)

Rainfall(Philippine, heavy)

Water-in-soil(Philippine, high)

(Philippine -> Vietnam)

determines determines ?

Rainfall(Vietnam, heavy)

Water-in-soil(Vietnam, high)(Philippine -> Vietnam)

Analogical reasoning based on determinations

What is the difference between a determination rule and a deductive rule?

Page 40: 09 analogical-learning

40 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Discussion

A determination rule is different from a deductive rule.

The form of a deductive rule is:U(x1,...,xn,y1,...,ym) --> V(x1,...,xn,y1,...,ym)

That is, the variables which appear in the left hand side of a rule also appear in the right hand side. Therefore, if we know that 'U(a1,...,an,b1,...,bm)' is true, we could apply modus ponens to infer that 'V(a1,...,an,b1,...,bm)' is also true.

This type of reasoning is not possible in the case of a determinationU(x1,...,xn,y1,...,ym) >- V(x1,...,xn,z1,...,zp)

because we do not know the values of the variables z1,...,zp.

In order to apply a determination rule, one would need a source entity, as will be illustrated in the following.

Page 41: 09 analogical-learning

41 2002, G.Tecuci, Learning Agents Laboratory

The basic procedure for answering the query V(T, ?z) by analogy:

1. Find a determination such that U(?x, ?y) >- V(?x, ?z)(i.e. decide which determinations could be relevant for T: U(T, ?y) >- V(T, ?z))

2. Find 'a' such that U(T, a) (i.e. find how the facts are instantiated in the target)

3. Find a source S such that U(S, a) (i.e. find a suitable source)

4. Find 'b' such that V(S, b) (i.e. find the answer to the query from the source: U(S, a) >- V(S, b))

5. Return 'b' as the solution to the query(U(T, a) >- V(T, b))

U(S,a) U(T,a)

V(S,b) V(T,?z)b

Analogy based on determinations: Method

Page 42: 09 analogical-learning

42 2002, G.Tecuci, Learning Agents Laboratory

Let us consider the following targetNationality (Jack, UK), Male(Jack), Height(Jack, 6'), ...

and the problem of answering the following question by analogy What is the native language of Jack ? (i.e. Native-language(Jack, ?z))

1. Find a determination such that U(x, y) >- Native-language(x, z)Such a determination is: Nationality (x, y) >- Native-language(x, z)

2. Find 'a' such that Nationality (Jack, a)Nationality (Jack, UK) a = UK

3. Find a source S such that Nationality (S, UK)Nationality (Jill, UK), Female(Jill) , Height(Jill, 5'10"),Native-Language(Jill, English)S = Jill

4. Find 'b' in S such that NativeLanguage(Jill, b)Native-Language(Jill, English) b = English

5. Return 'English' as the solution to the queryNative-language(Jack, English)

Analogy based on determinations: Illustration

Page 43: 09 analogical-learning

43 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Discussion

Consider the determination rule:U(x1,...,xn,y1,...,ym) >- V(x1,...,xn,z1,...,zp)

Should U and V be terms or could they be arbitrary logical expressions? Why?

What if we cannot find a source S for applying the determination?

Page 44: 09 analogical-learning

44 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Discussion

U and V may be an logical expressions.

Example:

The rainfall of a flat area determines the quantity of water in the soil of the areaRainfall(x, y) & Terrain(x, flat) --> Water-in-soil(x, z)

Rainfall(Philippines, heavy), Terrain(Philippines, flat), Water-supply(Philippines, high)Rainfall(Vietnam, heavy), Terrain(Vietnam, flat)Water-in-soil(Vietnam, ?t)

Rainfall(Philippines, heavy) Terrain(Philippines, flat)

Water-in-soil(Philippines, high)

(Philippines <- Vietnam)

determines determines ?

Rainfall(Vietnam, heavy) Terrain(Vietnam, flat)

Water-in-soil(Vietnam, high)(Philippines <- Vietnam)

Page 45: 09 analogical-learning

45 2002, G.Tecuci, Learning Agents Laboratory

Determinations: Discussion

What if we cannot find a source S for applying the determination?

Sometimes there is no source S such that U(S, a) is true, but one may find S' such that U(S', a') is true. In such a situation one needs a way to decide whether a' and a are similar enough to infer V(T, b). Therefore, even in the case of determinations one may need a matching function.

Example: Latitude of an area determines the climate of the areaLatitude(x, y) --> Climate(x, z)Latitude(Romania, 45°), Climate(Romania, temperate)Latitude(France, 47°)

Latitude(Romania, 45°)

Climate(Romania, temperate)

(Romania <- France)

determines determines ?

(Romania <- France) Climate(France, temperate)

Latitude(France, 47°)

Page 46: 09 analogical-learning

46 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 47: 09 analogical-learning

47 2002, G.Tecuci, Learning Agents Laboratory

Problem solving by analogy

Analogy means deriving new knowledge about an input entity by transferring it from a known similar entity.

How could we define problem solving by analogy?

Page 48: 09 analogical-learning

48 2002, G.Tecuci, Learning Agents Laboratory

Problem solving by analogy: definition

Problem solving by analogy is the process of transferring knowledge from past problem-solving episodes to new problems that share significant aspects with corresponding past experience and using the transferred knowledge to construct solutions to the new problems.

What could be the overall structure of a problem solving by analogy method?

Page 49: 09 analogical-learning

49 2002, G.Tecuci, Learning Agents Laboratory

The problem solving by analogy method

Let P be a problem to solve.

First, look into the knowledge base for a previous problem solving episode which shares significant aspects with the problem to solve.

Next transform the past episode to obtain a solution to the current problem.

What it means for problems to share significant aspects?How is the past problem solving episode transformed so as to obtain the solution to the current problem?

What questions need to be answered to develop such a method?

Page 50: 09 analogical-learning

50 2002, G.Tecuci, Learning Agents Laboratory

The derivational analogy method (Carbonell)

•Two problems share significant aspects if they match within a certain threshold, according to a given similarity metric. •The solution to the retrieved problem is perturbed incrementally until it satisfies the requirements of the new problem.

PreviouslySolved

Problem

NewProblem

Solutionto Old

Problem

Solutionto NewProblem

PartialMapping

TransformationProcess

Page 51: 09 analogical-learning

51 2002, G.Tecuci, Learning Agents Laboratory

The derivational analogy method: illustration

GIVEN: PROVE: AC = BD

A

BC

D

AB = CD

AB + BC = BC + CDAC = BD

AB = CD GIVEN: PROVE: <BAD = <CAE

<BAC = <DAE

B

A

C

D

E

BC = BC<BAC = <DAE

<BAC + <CAD= <CAD + <DAE<BAD = <CAE

<CAD = <CAD

AB <- <BAC CD <- <DAE AC <- <BAD BD <- <CAE )

= (

Page 52: 09 analogical-learning

52 2002, G.Tecuci, Learning Agents Laboratory

The derivational analogy method: discussion

How does analogy facilitate the problem solving process?

How does the derivational analogy method relates to the generally accepted idea that the relations which are usually imported by analogy from a source concept S to the target concept T are those belonging to causal networks?

Page 53: 09 analogical-learning

53 2002, G.Tecuci, Learning Agents Laboratory

The derivational analogy method: discussion

How does this method relates to the generally accepted idea that the relations which are usually imported by analogy from a source concept S to the target concept T are those belonging to causal networks?

Intuition: The relation between a problem and its solution is a kind of cause-effect relationship.

Fermat’s last theorem: There is no integer solutions of xn + yn = zn for n>2 Previously solved problem: Find integer solutions of the problem x2 + y2 = z2

Problem: Find integer solutions of the problem x3 + y3 = z3

Consider the following problem solving situation:

What does this example suggests?

Page 54: 09 analogical-learning

54 2002, G.Tecuci, Learning Agents Laboratory

Discussion

Except for the trivial problems, a solution does not emerge immediately from the problem formulation, as would be the case in a cause-effect relation.

What other relation from the problem solving process might be closer to a cause-effect relation?

Page 55: 09 analogical-learning

55 2002, G.Tecuci, Learning Agents Laboratory

What other relation from the problem solving process might be closer to a cause-effect relation?

The relation between a problem and its derivation trace (i.e. solution process).

What is transferred from a past problem solving episode is not a problem solution but the problem solving process itself, what questions have been asked, what factors have been considered, etc. One would try to repeat the same process in the context of the new problem.

With this interpretation we retrieve the derivational analogy method:

Discussion

Page 56: 09 analogical-learning

56 2002, G.Tecuci, Learning Agents Laboratory

The transformational analogy method (Carbonell)

Two problems are considered to share significant aspects if their initial analysis yields the same reasoning steps, that is, if the initial segments of their respective derivations start by considering the same issues and making the same decisions;

The derivation of the solved problem may therefore be transferred to the new problem by reconsidering the old decisions in the light of the new problem situation, preserving those that apply, and replacing or modifying those whose supports are no longer valid in the new situation.

Derivational analogy gives better results than transformational analogy. However, it has the disadvantage to manipulate complex structures representing derivational traces.

Page 57: 09 analogical-learning

57 2002, G.Tecuci, Learning Agents Laboratory

Overview

Learning by analogy: definition

Design issues

The structure mapping theory

Problem solving by analogy

Determinations

Recommended reading

Exercises

Page 58: 09 analogical-learning

58 2002, G.Tecuci, Learning Agents Laboratory

1. Compare explanation-based learning, empirical inductive learning, and learning by analogy from the point of view of input information, background knowledge needed, and outcome of learning.

2. Define learning by analogy and give an example of analogy.

3. Describe the four stages of learning by analogy.

4. Illustrate learning by analogy with the help of the following example:

Exercises

sun

planet

yellow

mass

mass

temperature

greater

color

revolves- around

attractsTsun

Tplanet

Msun

Mplanet

causes

temperature

greater

mass

mass

attracts

Mnucleus

greater

nucleus

electron

Melectron

Page 59: 09 analogical-learning

59 2002, G.Tecuci, Learning Agents Laboratory

Let us consider a learning by analogy system having the following background knowledge:

Facts:Economy-type (Germany, highly-industrial), Location(Germany, Europe)Population(Germany, 70), …Economy-type (Vietnam, agricultural), Location(Vietnam, Asia)Population(Vietnam, 70), Economic-state(Vietnam, poor), …Economy-type (Japan, highly-industrial), Location(Japan, Asia),Population(Japan, 100), Economic-state(Japan, excellent), …

Determination:Economy-type (x, y) >- Economic-state (x, z)(the economy-type of a country determines the economic state of the country)

Write a detailed trace of the reasoning of the system for answering the following question:

What is the economic state of Germany ? (i.e. Economic-state (Germany, ?z))

Exercise

Page 60: 09 analogical-learning

60 2002, G.Tecuci, Learning Agents Laboratory

Provide an example of a successful application of the transformational analysis method.

Provide an example where the transformational analysis method does not apply, but the derivational analogy method does apply.

What is the difference between a determination rule and a deductive rule?Illustrate the difference with an example.

Exercises

Page 61: 09 analogical-learning

61 2002, G.Tecuci, Learning Agents Laboratory

Recommended readingGentner D., Holyoak K.J., Kokinov B.N. (eds.), The Analogical Mind: Perspectives from Cognitive Science, The MIT Press, 2001.

Carbonell J.G., Learning by analogy: formulating and generalizing plans from past experience, Machine learning I, 1983.

Carbonell J.G., Derivational analogy: a theory of reconstructive problem solving and expertise acquisition, in Shavlik J. and Dietterich T. (eds), Readings in Machine Learning, Morgan Kaufmann, 1990. Also in Readings in Machine Learning and Knowledge Acquisition.

Davies T.R., Russell S.J., A logical approach to reasoning by analogy, in Shavlik J. and Dietterich T. (eds), Readings in Machine Learning, Morgan Kaufmann, 1990.

Gentner D., The mechanisms of analogical reasoning, in J.W.Shavlik, T.G.Dietterich (eds), Readings in Machine Learning, Morgan Kaufmann, 1990.

Winston P.H., Learning and reasoning by analogy, Communications of the ACM, 23, pp.689-703, 1980.

Forbus K.D., Exploring Analogy in the Large, in Gentner D., Holyoak K.J., Kokinov B.N. (eds.), The Analogical Mind, 2001

Tecuci, Building Intelligent Agents: An Apprenticeship Multistrategy LearningTheory, Methodology, Tool and Case Studies, Academic Press, 1998, pp: 101-108.