Cascade Principles, Bayes Rule and Wisdom of the

Cascade Principles, Bayes Rule and Wisdom of the

Cascade Principles, Bayes Rule and Wisdom of the Crowds Lecture 6 (Largely drawn from Kleinberg book) Following the crowd We are often influenced by others Opinions Political positions Fashion Technologies to use

Why do we sometimes imitate the choices of others even if information suggests otherwise? Why do you smoke? Why did you vote for a particular party? Why did you guess a particular color? Following the crowd It could be rational to do so:

You pick some restaurant A in an unfamiliar part of town Nobody there, but many others sitting at a restaurant B Maybe they have more information than you! You join them regardless of your own private information This is called herding, or an information cascade Following the crowd Milgram,

Bickman, Berkowitz in1960 x number of people stare up How many passers by will also look up? Increasing social force for conformity? Or expect those looking up to have more information? Information cascades partly explain many imitations in social settings Fashion, fads, voting for popular candidates Self-reinforcing success of books on highseller lists

Herding There is a decision to be made People make the decision sequentially Each person has some private information that helps guide the decision You cant directly observe the private information of others Can make inferences about their private information

Rational reasons Informational effects Wisdom of the crowds Direct-benefit effects Different set of reasons for imitation Maybe aligning yourself with others directly benefits you Consider the first fax machine Operating systems

Facebook We will consider the first one today Modeling information cascades Pr[A] where A is some event What is the probability this is the better restaurant? Pr[A | B] where A and B are events What is the probability this is the

better restaurant, given the reviews I read? Probability of A given B. Modeling information cascades Def: So: Notation P[A] = prior probability of A P[A | B] = posterior probability of A given B

Using Bayes rule Applies when assessing the probability that a particular choice is the best one, given the event that we received certain private information Lets take an example Bayes rule, example Crime

in a city involving a taxi 80% of taxis are black 20% of taxis are yellow Eyewitness testimony 80% accurate What is the probability that a taxi is yellow if the witness said it was? True = actual color of vehicle Report = color stated by witness

Want: Pr[true = Y | report = Y] Bayes rule, example We can compute this: If report is yellow, two possibilities: Cab is truly yellow

Cab is actually black So Bayes rule, example Putting it together Conclusion: Even though witness said taxi was yellow, it is equally likely to be truly yellow or black! Second example

Spam filtering Suppose: 40% of your e-mail is spam 1% of spam has the phrase check this out 0.4% of non-spam contain the phrase Apply Bayes rule!

Second example Numerator is easy 0.4 * 0.01 = 0.004 Denominator: So General cascade model Group of people sequentially making decisions

Choice between accepting or rejecting some option Wear a new fashion Buy new technology (I) State of the world Randomly in one of two states: The option is a good idea (G) The option is a bad idea (B) General cascade model Everyone

knows probability of the state World is in state G with probability p World is in state B with probability 1p (II) Payoffs Reject: payoff of 0 Accept a good option: vg > 0 Accept a bad option: vb < 0 General cascade model

(III) Signals Model the effect of private information High signal (H): Suggests that accepting is a good idea Low signal (L): Suggests that accepting is a bad idea Make this precise: General cascade model

Three main ingredients (I) State of the world (II) Payoffs (III) Signals Herding fits this framework General cascade model Consider an individual

Suppose he only uses private information If he gets high signal: Shifts To: What is this probability? General cascade model So

high signal = should accept Makes intuitive sense since option more likely to to be good than bad Analogous for low signal (should reject) What about multiple signals? Information from all the other people Can use Bayes rule for this

Suppose I see a sequence S with a high signals and b low ones General cascade model So what does a person decide given a sequence S? Want the following facts Accept if more high signals than low ones Lets derive this

General cascade model How does this compare to p? General cascade model Suppose we changed the term Whole expression becomes p Does this replacement make the

denominator smaller or larger? Herding experiment Using the model we can derive: People >3 will ignore own signal Cascades - lessons Cascades can be wrong

Accepting an option may be a bad idea But if first two people get high signals cascade of acceptances Cascades can be based on very little information People ignore private information once cascade starts Cascades are fragile

Adding even a little bit more information can stop even a longrunning cascade

Recently Viewed Presentations

  • Adaptations of a Megalodon shark

    Adaptations of a Megalodon shark

    Adaptations of a Megalodon shark. By:Brayden Wilkerson. Jaw size. I am 50feet when im open but when im closed im only 30feet. It helps me it huge animals like a blue whale. Body size. The body size was 70 feet...
  • CS267: Introduction - UBC

    CS267: Introduction - UBC

    Device (a.k.a. global) slow - 400-600 cycles access latency large - 256MB - 1GB Shared fast - 4 cycles access latency small - 16KB Texture - read only Constant - read only TTotal = 1 TPreprocesing 1 2 + TDataHtoG...
  • Partial Orderings Sometimes, relations define an order on

    Partial Orderings Sometimes, relations define an order on

    Sometimes, relations define an order on the elements in a set. Definition: A relation R on a set S is called a . partial ordering. or . partial order. if it is reflexive, antisymmetric, and transitive. A set S together...
  • OAIS in the Lifecycle Management of Records

    OAIS in the Lifecycle Management of Records

    "A Semi-Markov Survivability Evaluation Model for Intrusion Tolerant Database Systems". 2010 International Conference on Availability, Reliability and Security. ... Markov property: next state depends only on current state. p ij = p(X n+1 = j | X n = i,...
  • BMS 631 - LECTURE 1 Flow Cytometry: Theory J.Paul Robinson ...

    BMS 631 - LECTURE 1 Flow Cytometry: Theory J.Paul Robinson ...

    Times New Roman Comic Sans MS Symbol Arial Default Design Adobe Photoshop Image BMS 602/631 - LECTURE 8 Flow Cytometry: Theory J. Paul Robinson Professor of Immunopharmacology& Biomedical Engineering Purdue University Detectors Silicon photodiodes PMT Signal Detection - PMTs Photomultiplier...
  • Contract Support Costs Self-determination Agreements

    Contract Support Costs Self-determination Agreements

    Contract Support Cost Policy. Page 2 . 3. Policy. It is the policy to provide for a uniform and equitable system of distributing CSC funds to new and existing P.L.93-638 compacts and contracts, and to preserve and support each awardee's...
  • Scoliosis - ТДМУ

    Scoliosis - ТДМУ

    Scoliosis * * Answer 3 The answer is B: Although surgery for scoliosis is generally not recommended without marked curvature, well-conducted outcomes studies with patients who have had surgery have not been completed. Symptoms of back pain do not appear...
  • EmployeeChoice Helping employees stay healthy all year long

    EmployeeChoice Helping employees stay healthy all year long

    (.90 RAF is the best RAF available) 6% Life savings plus 6% Dental savings - Employers purchasing both $25,000 or more in Life and any of our insured Dental plans at the same time, will receive 6% savings on their...