Announcements

1. Announcements CS Ice Cream Social • 9/5 3:30-4:30, ECCR 265 • includes poster session, student group presentations

2. Concept Learning • Examples • Word meanings • Edible foods • Abstract structures (e.g., irony) glorch not glorch glorch not glorch

3. Supervised Approach To Concept Learning • Both positive and negative examples provided • Typical models (both in ML and Cog Sci) circa 2000 required both positive and negative examples

4. Contrast With Human Learning Abiliites • Learning from positive examples only • Learning from a small number of examples • E.g., word meanings • E.g., learning appropriate social behavior • E.g., instruction on some skill • What would it mean to learn from asmall number of positive examples? + + +

5. Tenenbaum (1999) • Two dimensional continuous feature space • Concepts defined by axis-parallel rectangles • e.g., feature dimensions • cholesterol level • insulin level • e.g., concept healthy

6. Learning Problem • Given a set of given a set of n examples,X = {x1, x2, x3, …, xn}, which are instances of the concept… • Will some unknown example Y also be an instance of the concept? • Problem of generalization + + 1 + 2 3

7. Hypothesis (Model) Space • H: all rectangles on the plane,parameterized by (l1, l2, s1, s2) • h: one particular hypothesis • Note: |H| = ∞ • Consider all hypotheses in parallel • In contrast to non-Bayesian approach of maintaining only the best hypothesisat any point in time.

8. Prediction Via Model Averaging Will some unknown input y be in the concept given examples X = {x1, x2, x3, …, xn}? Q: y is a positive example of the concept (T,F) • P(Q | X) = ⌠hp(Q & h | X) dh • P(Q & h | X) = p(Q | h, X) p(h | X) • P(Q | h, X) = P(Q | h) = 1 if y is in h • p(h | X) ~ P(X | h) p(h) Marginalization Chain rule Conditional independence and deterministic concepts Bayes rule likelihood prior

9. Priors and Likelihood Functions Priors, p(h) • Location invariant • Uninformative prior(prior depends only on area of rectangle) • Expected size prior • Likelihood function, p(X|h) • X = set of n examples • Size principle x

10. Expected size prior

11. Generalization Gradients • MIN: smallest hypothesis consistent with data • weak Bayes: instead of using size principle, assumes examples are produced by process independent of the true class Dark line =50% prob.

12. Experimental Design • Subjects shown n dots on screen that are “randomly chosen examples from some rectangle of healthy levels” • n drawn from{2, 3, 4, 6, 10, 50} • Dots varied in horizontal and vertical range • rdrawn from {.25, .5, 1, 2, 4, 8} units in a 24 unit window • Task • draw the ‘true’ rectangle around the dots

13. Experimental Results

14. Number Game • Experimenter picks integer arithmetic concept C • E.g., prime number • E.g., number between 10 and 20 • E.g., multiple of 5 • Experimenter presents positive examples drawn at random from C, say, in range [1, 100] • Participant asked whether some new test case belongs in C

15. Empirical Predictive Distributions

16. Hypothesis Space • Even numbers • Odd numbers • Squares • Multiples of n • Ends in n • Powers of n • All numbers • Intervals [n, m] for n>0, m<101 • Powers of 2, plus 37 • Powers of 2, except for 32

17. Observation = 16 • Likelihood function • Size principle • Prior • Intuition

18. Observation = 16 8 2 64 • Likelihood function • Size principle • Prior • Intuition

19. Posterior Distribution After Observing 16

20. Model Vs. Human Data MODEL HUMAN DATA

21. Summary of Tenenbaum (1999) • Method • Pick prior distribution (includes hypothesis space) • Pick likelihood function (size principle) • leads to predictions for generalization as a function of r (range) and n (number of examples) Claims people generalize optimally given assumptions about priors and likelihood Bayesian approach provides best description of how people generalize on rectangle task. Explains how people can learn from a small number of examples, and only positive examples.

22. Important Ideas in Bayesian Models Generative models • Likelihood function Consideration of multiple models in parallel • Potentially infinite model space Inference • prediction via model averaging • role of priors diminishes with amount of evidence Learning • trade off between model simplicity and fit to data  Bayesian Occam’s Razor

23. Ockham's Razor medieval philosopher and monk tool for cutting (metaphorical) If two hypotheses are equally consistent with the data, prefer the simpler one. Simplicity • can accommodate fewer observations • smoother • fewer parameters • restricts predictions more(“sharper” predictions) Examples 1st vs. 4th order polynomial small rectangle vs. large rectanglein Tenenbaum model

24. Motivating Ockham's Razor • Aesthetic considerations • A theory with mathematical beauty is more likely to be right (or believed) than an ugly one, given that both fit the same data. • Past empirical success of the principle • Coherent inference, as embodied by Bayesian reasoning, automatically incorporates Ockham's razor • Two theories H1 and H2 PRIORS LIKELIHOODS

25. Ockham's Razor with Priors Jeffreys (1939) probabililty text • more complex hypotheses should have lower priors Requires a numerical rule for assessing complexity • e.g., number of free parameters • e.g., Vapnik-Chervonenkis (VC) dimension

26. Subjective vs. Objective Priors subjective or informative prior • specific, definite information about a random variable objective or uninformative prior • vague, general information Philosophical arguments for certain priors as uninformative • Maximum entropy / least committment e.g., interval [a b]: uniform e.g., interval [0, ∞) with mean 1/λ: exponential distribution e.g., mean μ and std deviation σ: Gaussian • Independence of measurement scale e.g., Jeffrey’s prior 1/(θ(1-θ)) forθin [0,1]expresses same belief whether we talkabout θ or logθ

27. Ockham’s Razor Via Likelihoods • Coin flipping example • H1: coin has two headsH2: coin has a head and a tail • Consider 5 flips producing HHHHH • H1 could produce only this sequenceH2 could produce HHHHH, but also HHHHT, HHHTH, ... TTTTT • P(HHHHH | H1) = 1, P(HHHHH | H2) = 1/32 • H2 pays the price of having a lower likelihood via the fact it can accommodate a greater range of observations • H1 is more readily rejected by observations

28. Simple and Complex Hypotheses H2 H1

29. Bayes Factor • BIC is approximation to Bayes factor • A.k.a. likelihood ratio

30. Hypothesis Classes Varying In Complexity • E.g., 1st, 2nd, and 3d order polynomials • Hypothesis class is parameterized by w v

31. Rissanen (1976)Minimum Description Length • Prefer models that can communicate the data in the smallest number of bits. • The preferred hypothesis H for explaining data D minimizes: • (1) length of the description of the hypothesis • (2) length of the description of the data with the help of the chosen theory L: length

32. MDL & Bayes • L: some measure of length (complexity) • MDL: prefer hypothesis that min. L(H) + L(D|H) • Bayes rule implies MDL principle • P(H|D) = P(D|H)P(H) / P(D) • –log P(H|D) = –log P(D|H) – log P(H) + log P(D) = L(D|H) + L(H) + const

34. Relativity Example • Explain deviation in Mercury's orbit at perihelion with respect to prevailing theory • E: Einstein's theory α = true deviationF: fudged Newtonian theory a = observed deviation

35. Relativity Example (Continued) • Subjective Ockham's razor • result depends on one's belief about P(α|F) • Objective Ockham's razor • for Mercury example, RHS is 15.04 • Applies to generic situation