/
The “curse of dimensionality” The “curse of dimensionality”

The “curse of dimensionality” - PowerPoint Presentation

liane-varnes
liane-varnes . @liane-varnes
Follow
429 views
Uploaded On 2017-01-25

The “curse of dimensionality” - PPT Presentation

And the impossibility of utility maximisation From indifference Concept of subjective utility dates back at least to Aristotle made central tenet of economics and philosophy by Jeremy Bentham ID: 513919

utility 000 consumer combinations 000 utility combinations consumer indifference budget bananas biscuits commodities combination level line neoclassical numbers commodity

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "The “curse of dimensionality”" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

The “curse of dimensionality”

And the impossibility of utility maximisationSlide2

From indifference…

Concept of subjective utility dates back at least to Aristotle; made central tenet of economics and philosophy by Jeremy Bentham (

not

Adam Smith!)

Formalised into concept of “indifference curves” by neoclassical economists (see Chapter 2,

Debunking Economics

)

Many different combinations of goods capable of giving consumer same level of subjective utility

Line linking points akin to “isoquant” on map, linking points of equivalent altitude

Christened “indifference curves”:Slide3

Bananas

Biscuits

A

B

C

D

Combinations A and B give same level of satisfaction

Combination C gives higher level than A or B

Combination D gives lower level than A or B

From indifference…

Just one wee problem…

Indifference curves no more observable than angels dancing on heads of medieval pins…Slide4

To revealed preference…

Samuelson proposed solution: “revealed preference”

Indifference curves could be inferred from behaviour of consumer, provided 4 postulates of rationality met:

Completeness: consumer knows own subjective ranking of all combinations of goods

Transitivity: If combination A preferred to combination B, and B to C, then A preferred to C

Non-satiation: More is always preferred to less.

Convexity: additional utility a consumer gets from extra units of each commodity fallsPreferences and budget completely independentApplying this:Slide5

Bananas

Biscuits

A

To revealed preference…

All points in rectangle necessarily preferred to A (“non-satiation”)

Bananas

Biscuits

B

A

C

If A purchased when both B and C affordable, then A necessarily on higher indifference curve than B and C

Move budget line around, see responses, can eventually infer indifference map from observed behaviour

Appears scientific, but problems when verification attempted…Slide6

To experimental testing…

Bananas

Biscuits

In theory, consumer starts with

complete

preference set

Imposes budget line

Works out point of tangency

Buys this combination

In experiment, easy to decide whether someone is rational or irrational according to theory:

(Preferences and budget completely independent)Slide7

The neoclassical expectation…

Bananas

Biscuits

X Initial budget line

Consumer chooses A when A & B both affordable

A must lie on higher indifference curve

Rational consumer “should” always prefer A to B

But in experiments they don’t do this! Sometimes, they choose B instead of A

A

X

B

Budget Y: A no longer best, but “clearly” better than B

YSlide8

Consumers “irrational”

But in reality, economists definition of “rational” is irrational

Many weaknesses

but key one the unwitting assumption that consumers have near infinite information processing ability: the “curse of dimensionality”

issue easily illustrated by putting numbers on axes of economists’ graphs:

The neoclassical explanation…Slide9

Putting some numbers on the graph…

Bananas

Biscuits

0

1

2

3

4

5

6

7

8

9

10

0

1

2

3

4

5

6

7

8

9

10

121 combinations

Some you ignore

Others you can’t…

10 pairs

10 additions

10 comparisons

Easy!

But…Slide10

Putting some numbers on the graph…

Every additional commodity considered adds another dimension. With no more than 10 units of each:

2 commodities, ~100 combinations

(actually 121)

3 commodities, ~1,000 combinations

4 commodities, 10,000 combinations

30 commodities…

how many combinations?Slide11

Putting some numbers on the graph…

10,000,000,000,000,000,000,000,000,000,000!

If budget obviously ruled out 99.9% of these;

If each evaluation took 1 billionth of a second…

Process would complete after

32 billion years

Maximum (est.) age of universe 20 billion yearsIndividual would take 1.6 times age of known universe to make “utility maximising” choice of just 30 commoditiesMaximising utility in typical supermarket (1,000+ different items) doesn’t bear contemplationlet alone millions of products in modern economyInstead, intelligent partitioning of commodity space vital…Slide12

Utility maximising individuals?

Sources of partitioning

Culture

some commodities not even contemplated (sea slugs, anyone?)

Income & Needs

Fulfil hierarchy of wants from basic to ethereal (Maslow’s concepts)

Rolls Royce’s not part of “utility set” of poorBaked beans not part of “utility set” of richHabitBuy mainly what you bought yesterdayChange in consumptionEvolutionary rather than “rational” processSlide13

Utility maximising individuals?

Individual tastes no longer a “given” but vital economic issue

Explains individual partitioning of commodity space

Selling new products requires movement of this space

Marketing, advertising thus essential “economic” activities if new products are to be sold

Co-evolution of products and tastes an essential aspect of economic development

Consumers do not “utility maximise” but instead “satisfice” (as per Herbert Simon)Demand curve for individual consumer becomes meaningless, let alone neoclassical idea of market demand curve (see Debunking Economics, Chapter 2)