bayesian learning regression based learning n.
Download
Skip this Video
Loading SlideShow in 5 Seconds..
Bayesian Learning, Regression-based learning PowerPoint Presentation
Download Presentation
Bayesian Learning, Regression-based learning

Loading in 2 Seconds...

play fullscreen
1 / 52

Bayesian Learning, Regression-based learning - PowerPoint PPT Presentation


  • 225 Views
  • Uploaded on

Bayesian Learning, Regression-based learning. Overview. Bayesian Learning Full MAP learning Maximum Likelihood Learning Learning Bayesian Networks (Fully observable) Regression and Logistic Regression. Full Bayesian Learning.

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about 'Bayesian Learning, Regression-based learning' - cai


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
overview
Overview
  • Bayesian Learning
    • Full
    • MAP learning
    • Maximum Likelihood Learning
      • Learning Bayesian Networks (Fully observable)
  • Regression and Logistic Regression
full bayesian learning
Full Bayesian Learning
  • In Decision Trees (and in all non-Bayesian learning methods) the idea is always to find the best model that explains some observations
  • In contrast, full Bayesian learning sees learning as Bayesian updating of a probability distribution over the hypothesis space, given data
    • H is the hypothesis variable
    • Possible hypotheses (values of H) h1…, hn
    • P(H) = prior probability distribution over hypotesis space
  • jth observation dj gives the outcome of random variable Dj
    • training data d= d1,..,dk
full bayesian learning1
Full Bayesian Learning
  • Given the data so far, each hypothesis hihas a posterior probability:
    • P(hi |d) = αP(d| hi) P(hi) (Bayes theorem)
    • where P(d| hi) is called the likelihoodof the data under each hypothesis
  • Predictions over a new entity X are a weighted average over the prediction of each hypothesis:
    • P(X|d) =

= ∑iP(X, hi |d)

= ∑iP(X| hi,d) P(hi |d)

= ∑iP(X| hi) P(hi |d)

~ ∑iP(X| hi) P(d| hi) P(hi)

    • The weights are given by the data likelihood and prior of each h
  • No need to pick one best-guess hypothesis!

The data does not add anything to a prediction given an hp

example
Example
  • Suppose we have 5 types of candy bags
    • 10% are 100% cherry candies (h100, P(h100 )= 0.1)
    • 20% are 75% cherry + 25% lime candies (h75 , P(h75 )= 0.2)
    • 40% are 50% cherry + 50% lime candies (h50 , P(h50 )= 0.4)
    • 20% are 25% cherry + 75% lime candies (h25 , P(h25 )= 0.2)
    • 10% are 100% lime candies (h0,P(h100 )= 0.1)
  • The we observe candies drawn from some bag
  • Let’s call θ the parameter that defines the fraction of cherry candy in a bag, and hθthe corresponding hypothesis
  • Which of the five kinds of bag has generated my 10 observations? P(h θ |d).
  • What flavour will the next candy be? Prediction P(X|d)
example1
Example
  • If we re-wrap each candy and return it to the bag, our 10 observations are independent and identically distributed, i.i.d, so
    • P(d| hθ) = ∏j P(dj| hθ) for j=1,..,10
  • For a given hθ , the value of P(dj| hθ) is
    • P(dj = cherry| hθ) = θ; P(dj = lime|hθ) = (1-θ)
  • And given N observations, of which c are cherry and l = N-c lime
    • Binomial distribution: probability of # of successes in a sequence of N independent trials with binary outcome, each of which yields success with probability θ.
  • For instance, after observing 3 lime candies in a row:
    • P([lime, lime, lime| h 50) = 0.53 because the probability of seeing lime for each observation is 0.5 under this hypotheses
posterior probability of h

P(h100|d)

P(h75|d)

P(h50|d)

P(h25|d)

P(h0|d)

Posterior Probability of H

P(hi |d) = αP(d| hi) P(hi)

  • Initially, the hp with higher priors dominate (h50 with prior = 0.4)
  • As data comes in, the true hypothesis (h0 ) starts dominating, as the probability of seeing this data given the other hypotheses gets increasingly smaller
    • After seeing three lime candies in a row, the probability that the bag is the all-lime one starts taking off
prediction probability
Prediction Probability

∑iP(next candy is lime| hi) P(hi |d)

  • The probability that the next candy is lime increases with the probability that the bag is an all-lime one
overview1
Overview
  • Full Bayesian Learning
  • MAP learning
  • Maximun Likelihood Learning
  • Learning Bayesian Networks
    • Fully observable
    • With hidden (unobservable) variables
map approximation
MAP approximation
  • Full Bayesian learning seems like a very safe bet, but unfortunately it does not work well in practice
    • Summing over the hypothesis space is often intractable (e.g., 18,446,744,073,709,551,616 Boolean functions of 6 attributes)
  • Very common approximation: Maximum a posterior (MAP) learning:
    • Instead of doing prediction by considering all possible hypotheses , as in
      • P(X|d) = ∑iP(X| hi) P(hi |d)
    • Make predictions based on hMAP that maximises P(hi |d)
      • I.e., maximize P(d| hi) P(hi)
      • P(X|d)~ P(X|hMAP)
map approximation1

P(h100|d)

P(h75|d)

P(h50|d)

P(h25|d)

P(h0|d)

MAP approximation
  • Map is a good approximation when P(X |d) ≈ P(X| hMAP)
    • In our example, hMAPis the all-lime bag after only 3 candies, predicting that the next candy will be lime with p =1
    • the bayesian learner gave a prediction of 0.8, safer after seeing only 3 candies
slide12
Bias
  • As more data arrive, MAP and Bayesian prediction become closer, as MAP’s competing hypotheses become less likely
  • Often easier to find MAP (optimization problem) than deal with a large summation problem
  • P(H) plays an important role in both MAP and Full Bayesian Learning
    • Defines the learning bias, i.e. which hypotheses are favoured
  • Used to define a tradeoff between model complexity and its ability to fit the data
    • More complex models can explain the data better => higher P(d| hi) danger of overfitting
    • But simpler model have higher probability
    • I.e. common learning bias is to penalize complexity
overview2
Overview
  • Full Bayesian Learning
  • MAP learning
  • Maximun Likelihood Learning
  • Learning Bayesian Networks
    • Fully observable
    • With hidden (unobservable) variables
maximum likelihood ml learning
Maximum Likelihood (ML)Learning
  • Further simplification over full Bayesian and MAP learning
    • Assume uniform priors over the space of hypotheses
    • MAP learning (maximize P(d| hi) P(hi)) reduces to maximize P(d| hi)
  • When is ML appropriate?
maximum likelihood ml learning1
Maximum Likelihood (ML) Learning
  • Further simplification over Full Bayesian and MAP learning
    • Assume uniform prior over the space of hypotheses
    • MAP learning (maximize P(d| hi) P(hi)) reduces to maximize P(d| hi)
  • When is ML appropriate?
    • Used in statistics as the standard (non-bayesian) statistical learning method by those who distrust subjective nature of hypotheses priors
    • When the competing hypotheses are indeed equally likely (e.g. have same complexity)
    • With very large datasets, for which P(d| hi) tends to overcome the influence of P(hi))
overview3
Overview
  • Bayesian Learning
    • Full
    • MAP learning
    • Maximum Likelihood Learning
      • Learning Bayesian Networks (Fully observable)
learning bnets complete data
Learning BNets: Complete Data
  • We will start by applying ML to the simplest type of BNets learning:
    • known structure
    • Data containing observations for all variables
      • All variables are observable, no missing data
  • The only thing that we need to learn are the network’s parameters
ml learning example
ML learning: example
  • Back to the candy example:
    • New candy manufacturer that does not provide data on the probability of different types of bags, i.e. fraction θof cherry candy
    • Any θ is possible: continuum of hypotheses hθ
    • Reasonable to assume that all θ are equally likely (we have no evidence of the contrary): uniform distribution P(hθ)
    • θ is a parameter for this simple family of models, that we need to learn
  • Simple network to represent this problem
    • Flavor represents the event of drawing a cherry vs. lime candy from the bag
    • P(F=cherry), or P(cherry) for brevity, is equivalent to the fraction θ of cherry candies in the bag
  • We want to infer θby unwrapping N candies from the bag
ml learning example cont d
ML learning: example (cont’d)
  • Unwrap N candies, c cherries and l = N-c lime (and return each candy in the bag after observing flavor)
  • As we saw earlier, this is described by a binomial distribution
    • P(d| h θ) = ∏j P(dj| h θ) = θc (1- θ) l
  • With ML we want to find θthat maximizes this expression, or equivalently its log likelihood (L)
    • L(P(d| h θ)
    • = log (∏j P(dj| h θ))
    • = log (θc (1- θ) l)
    • = clogθ + l log(1- θ)
ml learning example cont d1
ML learning: example (cont’d)
  • To maximise, we differentiate L(P(d| h θ) with respect to θ and set the result to 0

the proportion of cherries in the bag is equal to the proportion (frequency) of in cherries in the data

  • Doing the math gives
general ml procedure
General ML procedure
  • Express the likelihood of the data as a function of the parameters to be learned
  • Take the derivative of the log likelihood with respect of each parameter
  • Find the parameter value that makes the derivative equal to 0
  • The last step can be computationally very expensive in real-world learning tasks

There is one more example for you to look at

in the next few slides

another example
Another example
  • The manufacturer choses the color of the wrapper probabilistically for each candy based on flavor, following an unknown distribution
    • If the flavour is cherry, it chooses a red wrapper with probability θ1
    • If the flavour is lime, it chooses a red wrapper with probability θ2
  • The Bayesian network for this problem includes 3 parameters to be learned
    • θθ1θ2
another example1
Another example
  • The manufacturer choses the color of the wrapper probabilistically for each candy based on flavor, following an unknown distribution
    • If the flavour is cherry, it chooses a red wrapper with probability θ1
    • If the flavour is lime, it chooses a red wrapper with probability θ2
  • The Bayesian network for this problem includes 3 parameters to be learned
    • θθ1θ2
another example cont d
Another example (cont’d)
  • P( W=green, F = cherry| hθθ1θ2) =(*)

= P( W=green|F = cherry, hθθ1θ2) P( F = cherry| hθθ1θ2)

= θ (1-θ1)

  • We unwrap N candies
    • c are cherry and l are lime
    • rc cherry with red wrapper, gc cherry with green wrapper
    • rllime with red wrapper, g llime with green wrapper
    • every trial is a combination of wrapper and candy flavor similar to event (*) above, so
  • P(d| hθθ1θ2)

= ∏j P(dj| hθθ1θ2)

    • = θc(1-θ) l (θ1) rc(1-θ1) gc(θ2) r l (1-θ2) g l
another example cont d1
Another example (cont’d)
  • I want to maximize the log of this expression
    • clogθ + l log(1- θ) + rclog θ1 + gclog(1- θ1) + rllog θ2 + g llog(1- θ2)
  • Take derivative with respect of each of θ, θ1 ,θ2
    • The terms not containing the derivation variable disappear

Frequencies again!

ml parameter learning in bayesian nets
ML parameter learning in Bayesian nets
  • With complete data and ML approach:
    • Parameters learning decomposes into a separate learning problem for each parameter (CPT), because of the log likelihood step
    • Each parameter is given by the frequency of the desired child value in the presence of the relevant parents values
    • P(Y=yi|X=xj) =
  • Frequencies are used to learn the relevant probabilities (e.g. transition and observation models), in HMM and more complex bayesian networks
    • See C&G island exercise and papers in presentation list
    • ML is the theoretical justification of this approach
very popular application
Very Popular Application

C

  • Naïve Bayes models: very simple Bayesian networks for classification
    • Class variable (to be predicted) is the root node
    • Attribute variables Xi (observations) are the leaves
  • Naïve because it assumes that the attributes are conditionally independent of each other given the class
  • Deterministic prediction can be obtained by picking the most likely class
  • Scales up really well: with nboolean attributes we just need…….

Xi

X1

X2

very popular application1
Very Popular Application

C

  • Naïve Bayes models: very simple Bayesian networks for classification
    • Class variable (to be predicted) is the root node
    • Attribute variables Xi (observations) are the leaves
  • Naïve because it assumes that the attributes are conditionally independent of each other given the class
  • Deterministic prediction can be obtained by picking the most likely class
  • Scales up really well: with nboolean attributes we just need 2n+1 parameters

Xi

X1

X2

example2
Example
  • Naïve Classifier for the newsgroup reading example
example3
Example
  • Naïve Classifier for the newsgroup reading example
problem with ml parameter learning
Problem with ML parameter learning
  • With small datasets, some of the frequencies may be 0 just because we have not observed the relevant data
  • Generates very strong incorrect predictions:
    • Simple common fix: initialize the count of every relevant event to 1 before counting the observations
    • For more sophisticated strategies see textbook (p. 296)
probability from experts
Probability from Experts
  • As we mentioned in previous lectures, an alternative to learning probabilities from data is to get them from experts
  • Problems
    • Experts may be reluctant to commit to specific probabilities that cannot be refined
    • How to represent the confidence in a given estimate
    • Getting the experts and their time in the first place
  • One promising approach is to leverage both sources when they are available
    • Get initial estimates from experts
    • Refine them with data
combining experts and data
Combining Experts and Data
  • Get the expert to express her belief on event A as the pair
    • <n,m>
    • i.e. how many observations of A they have seen (or expect to see) in m trials
  • Combine the pair with actual data
    • If A is observed, increment both n and m
    • If ⌐A is observed, increment m alone
  • The absolute values in the pair can be used to express the expert’s level of confidence in her estimate
    • Small values (e.g., <2,3>) represent low confidence
    • The larger the values, the higher the confidence

WHY?

combining experts and data1
Combining Experts and Data
  • Get the expert to express her belief on event A as the pair
    • <n,m>
    • i.e. how many observations of A they have seen (or expect to see) in m trials
  • Combine the pair with actual data
    • If A is observed, increment both n and m
    • If ⌐A is observed, increment m alone
  • The absolute values in the pair can be used to express the expert’s level of confidence in her estimate
    • Small values (e.g., <2,3>) represent low confidence, as they are quickly dominated by data
    • The larger the values, the higher the confidence as it takes more and more data to dominate the initial estimate (e.g. <2000, 3000>)
overview4
Overview
  • Bayesian Learning
    • Full
    • MAP learning
    • Maximum Likelihood Learning
      • Learning Bayesian Networks (Fully observable)
  • Regression and Logistic Regression
one more set of techniques fpr supervised learning
One more set of techniques fpr supervised learning
  • Naïve Bayes and Decision Trees (in their basic incarnation) provide classification in terms of discrete labels y.
  • y = f(attr1, attr2, …. attrn)
  • We will briefly see two techniques that go beyond
    • Regression: y is continuous
    • Logistic Regression: y is continuous but reduced to a binary classification
linear regression
Linear Regression

problem of fitting a linear function to a set of training examples: input/output pairs with numeric values

hw(x) = w1x + w0

regression
Regression
  • Linear regression: problem of fitting a linear function hw(x) to a set of training examples (xj,yj): input/output pairs with numeric values
    • y = m x + b
    • hw(x) = w1x + w0
  • Find best values for parameters that
    • “maximize goodness of fit” or “minimize loss”
    • Then most probable values of parameters found by minimizing squared-error loss:Loss(hw) = Σj(yj – hw(xj))2
regression minimizing loss
Regression: Minimizing Loss

Choose weights to

minimize

sum of squared errors

regression minimizing loss1
Regression: Minimizing Loss

y = w1x + w0

  • Loss(hw) = Σj(yj – hw(xj))2

Algebra givesan exact solution to

the minimization

problem

multivariate regression
Multivariate Regression
  • Suppose we have features X1,...,Xn. A linear function of these features is a function of the form

fw(X1,...,Xn) = w0+w1 × X1 + ...+ wn × Xn ,

  • Given a set E of examples, where each example e∈E has
    • values xi for feature Xi
    • observed value oe.
  • The predicted value is thus

hwe=w0+w1× x1 + ...+ wn× xn =∑i=0n wi× xi,

where x0is defined to be 1.

  • The sum-of-squares error on examples E for target Y is

ErrorE(w) = ∑e∈E(oe. -hwe)2

    • In this linear case, as for the univariate version, the weights that minimize the error can be computed analytically, by equating the derivatives in each wi to 0
squashed linear functions for classification
Squashed linear functions for classification
  • We can use a linear function for discreteclassification
  • Let’s consider binary classification task , where the domain of the target variable is {0,1}.
  • For classification, use a squashed linear function of the form
  • fw(X1,...,Xn) = f( w0+w1 × X1 + ...+ wn × Xn) ,
  • where f is an activation function, from real numbers into [0,1].
activation functions
Activation Functions
  • Commonly used ones:
    • f (w0 + w1*X1+ w2*X2+….+ wk*Xk)
  • Examples:

f(in) = 1 when in > t

= 0 otherwise

step function
Step Function
  • A step function represents a linear classifier with hard threshold
    • implements a linear decision boundary (aka linear separator) that separates the classes involved in the classification

Decision boundary

  • A step function was the basis for the perceptron, one of the first methods supervised classification and the basis for Neural Networks
  • Can’t use derivative to compute weights that minimize loss because step function is not derivable
expressiveness of step function
Expressiveness of Step Function
  • Step function can only represent linearly separable functions .

Linearly

separable

Non-Linearly

separable

Linearly

separable

Linearly

separable

= FALSE

Majority (I1,I2,I3)

= TRUE

logistic regression
Logistic Regression
  • Classification that uses a sigmoid function as activation function
  • Learning with logistic regression: still want to minimize the sum of squared errors over training set of examples E
  • ErrE(w) = ∑e∈E(oe- hwe)2
  • = ∑e∈E(oe - f(∑i wi× xi))2.
  • Because f(x) is a non-linear function, hard to find solution analitically:
      • Iterative computation of weights using Gradient Descent
gradient descent search
Gradient Descent Search
  • Look at the partial derivative of the error term on each example ewith respect to each weight wj ,
  • Alter each weight in an amount proportional to the slope in that direction
  • Repeat until termination condition is met (error is “small enough”)
  •  is a constant called the learning rate that determines how fast the algorithm moves toward the minimum of the error function
gradient descent search1
Gradient Descent Search
  • Search toward the minimum of the surface that describes the sum-of-squares errors as a function of all the weights in the logistic regression

each set of weights defines a point on the error surface

Given a point on the surface, look at the slope of the surface along the axis formed by each weight

partial derivative of the surface Err with respect to each weight wj

weight update for logistic regression
Weight Update for Logistic Regression
  • The predicted value for example eis
  • The actual observed value is oe
  • Partial derivative with respect to weight wj
weight update for logistic regression1

=

-

f

'

(

x

)

f

(

x

)(

1

f

(

x

))

chain rule

Weight Updatefor Logistic Regression
  • From this result and
  • each weight wj in gradient descent for Logistic Regression is updated as follows
  • The sigmoid activation function is used because its derivative is easy to compute analytically
expressiveness of logistic regression
Expressiveness of Logistic Regression
  • Sigmoid activation function has similar limitations in expressiveness as step function
  • Output of a two-input logistic regression
  • Adjusting weights changes the orientation, location and steepness of cliff, but still can’t learn functions like XOR
learning goals for bayesian learning and regression based learning
Learning Goals for Bayesian learning and regression-based learning
  • For each of Bayesian learning, MAP and ML learning
    • Define how it works
    • apply it to compute P(h|data) and P(X|data)
  • Explain how parameters can be learned from data in fully observable Bayesian networks.
  • Compute the parameters for a specific network
  • Explain how parameters can be learned in partially observable Bayesian networks (see slides from next class)
  • .