42

Heuristics g2

Embed Size (px)

Citation preview

Page 1: Heuristics g2
Page 2: Heuristics g2

R.SalimiN.Mohammadi

Page 3: Heuristics g2

CONTENT

Introduction

Heuristic Types

Gaze heuristic

Recognition heuristic

Social heuristic

Heuristics Based on Reasons

Take the best heuristic

Tallying heuristic

Hindsight bias

Conclusion

Page 4: Heuristics g2

Introduaction

Page 5: Heuristics g2

WHAT to DO

?

Page 6: Heuristics g2

What is Unbounded Rationality ?

Unbounded Rationality

infinite amount of time

all relevant informationIn models of Bounded

Rationality(Simon’s term):

how do humans who have little time & knowledge

behave?

Page 7: Heuristics g2

3 Programs associated with bounded rationality

1

• The study of optimization under constraints.

2• The study of cognitive illusion.

3

•The study of fast & frugal heuristic.

Page 8: Heuristics g2

Heuristic is a rule:

A rule is not necessarily a heuristic, unless it embodies 3 qualities

Heuristics Versus Rules

Page 9: Heuristics g2

1-Heuristics exploit evolved capacities

2-Heuristics exploit structures of environments

3- heuristics are distinct from “as if” optimization models

Page 10: Heuristics g2

Simon’s vision of bounded rationality

Simon: Human rational behavior is

structure of

task

environment

capabilities of the actor

Page 11: Heuristics g2

Heuristic Types

Gaze Heuristic

Page 12: Heuristics g2

Gaze heuristicPRABOBOLAInitial DistanceInitial VelocityProjection AngleProjection SpeedDirection of The Wind

No such robots exist yet!

Page 13: Heuristics g2

RUNNING SPEED

ANGLE OF GAZE REMAINS CONSTANT

NOT COMPUTE THE POINT AT WHICH THE BALL WILL LAND, BUT HE WILL BE THERE

Gaze heuristic

FAST and FRUGAL A REAL EXPERINCE . . .

Page 14: Heuristics g2

Heuristic Type

Recognition Heuristics

Page 15: Heuristics g2

TWO CITIES IN

UNITED STATES

San DiegoSan Antonio

Page 16: Heuristics g2

GERMAN

Which city has

more inhabitant?

33 %

66 %

SAN ANTONIO

SAN DIEGO 100 %

0 %

Recognition heuristic

AMERICANWHAT

HAPPENED …!?

Less Knowledge

More Correct Inference

Page 17: Heuristics g2

Ecological Rationality:

Correlation between recognition & criteria

Recognition validity:

α = R / ( R + W )

Page 18: Heuristics g2

recognition information

further knowledge

less-is-more effect

Page 19: Heuristics g2

Less-is-more in groups

Decision to go to museum

which museum?

Page 20: Heuristics g2

The less-is-more effect

)1(

)1(

2

1

)1(

)1)((

)1(

)(2

NN

nn

NN

nNnN

NN

nNnc

N: number of objects

n: number of recognized objects

α: recognition validity

β: knowledge validity

Page 21: Heuristics g2

The recognition heuristic will yield aless-is-more effect if α > β

Page 22: Heuristics g2

Heuristic Type

Social Heuristics

Page 23: Heuristics g2

“Do what the majority do” heuristic

Darwin

Decision to marry

Single or married

Page 24: Heuristics g2

Ecological Rationality of

do-what-the-majority do Heuristics

Similarenvironments

Stableenvironments

Noisy

environments

Page 25: Heuristics g2

Heuristic Type

Heuristics Based on Reasons

Page 26: Heuristics g2
Page 27: Heuristics g2

Heuristic Type

Take the Best &Tallying

Page 28: Heuristics g2

Take The Best heuristic (TTB)

ONE-REASON DECISION MAKINGSEARCH BY VALIDITY

Validities Cues School A School B

0.9 reputation + -

0.6 distance - +

0.8 students - +

ONE-REASON STOPPING RULE

Page 29: Heuristics g2

Validities Cues School A School B

0.9 reputation + -

0.6 distance - +

0.8 students - +

Random

search

Stop after

searching m

cues

Tallying rull

Tallying heuristic

Page 30: Heuristics g2

Ecological Rationality of Take The Best and Tallying

Take The BestNoncompensatory

Tallying Compensatory

Page 31: Heuristics g2

Hindsight Bias

Page 32: Heuristics g2

Hindsight Bias

Which has more cholesterol , cake or pie?

Cues Original Feedback Recall

Saturated fat (80%) cake? Pie "cake" cake> pie

Calories (70%) cake> pie

Protein (60%)

Choice cake cake

Confidence 70% 80%

Page 33: Heuristics g2

Which heuristic to use?

Page 34: Heuristics g2

Using heuristic unconsciously

adapt heuristics to changing environment with

feedback.

Page 35: Heuristics g2

Evaluating creditworthy

How people adapt their heuristics to

the structure of environment

Page 36: Heuristics g2

Robustness

• the ability to make predictions about the future or new events

fitness

• the ability to fit the past or already known data

Page 37: Heuristics g2

Simplicity can lead to higher predictive

accuracy

Page 38: Heuristics g2

The building Blocks of Heuristics

Page 39: Heuristics g2

Heuristics In Hospital

Page 40: Heuristics g2

CONCLUSION

Page 41: Heuristics g2

CONCLUSION

Page 42: Heuristics g2