Linear’Regression’ - Carnegie Mellon School of …mgormley/courses/10701-f16/...• However,...
Transcript of Linear’Regression’ - Carnegie Mellon School of …mgormley/courses/10701-f16/...• However,...
Linear Regression
1
Matt Gormley Lecture 4
September 19, 2016
School of Computer Science
Readings: Bishop, 3.1 Murphy, 7
10-‐701 Introduction to Machine Learning
Reminders
• Homework 1: – due 9/26/16
• Project Proposal: – due 10/3/16 – start early!
2
Outline • Linear Regression
– Simple example – Model
• Learning – Gradient Descent – SGD – Closed Form
• Advanced Topics – Geometric and Probabilistic Interpretation of LMS – L2 Regularization – L1 Regularization – Features – Locally-‐Weighted Linear Regression – Robust Regression
3
Outline • Linear Regression
– Simple example – Model
• Learning (aka. Least Squares) – Gradient Descent – SGD (aka. Least Mean Squares (LMS)) – Closed Form (aka. Normal Equations)
• Advanced Topics – Geometric and Probabilistic Interpretation of LMS – L2 Regularization (aka. Ridge Regression) – L1 Regularization (aka. LASSO) – Features (aka. non-‐linear basis functions) – Locally-‐Weighted Linear Regression – Robust Regression
4
• Our goal is to estimate w from a training data of <xi,yi> pairs
• Optimization goal: minimize squared error (least squares):
• Why least squares?
- minimizes squared distance between measurements and predicted line
- has a nice probabilistic interpretation
- the math is pretty
Linear regression in 1D
∑ −i
iiw wxy 2)(minargX
Y ε+= wxy
see HW
Slide courtesy of William Cohen
Solving Linear Regression in 1D
• To optimize – closed form:
• We just take the derivative w.r.t. to w and set to 0:
∂∂w
(yi −wxi )2
i∑ = 2 −xi (yi −wxi )
i∑ ⇒
2 xi (yi −wxi ) = 0i∑ ⇒
xiyi = wxi2
i∑
i∑ ⇒
w =xiyi
i∑
xi2
i∑
2 xiyii∑ − 2 wxixi
i∑ = 0
Slide courtesy of William Cohen
Linear regression in 1D
• Given an input x we would like to compute an output y
• In linear regression we assume that y and x are related with the following equation:
y = wx+ε where w is a parameter and ε
represents measurement error or other noise
X
Y
What we are trying to predict
Observed values
Slide courtesy of William Cohen
Regression example
• Generated: w=2 • Recovered: w=2.03 • Noise: std=1
Slide courtesy of William Cohen
Regression example
• Generated: w=2 • Recovered: w=2.05 • Noise: std=2
Slide courtesy of William Cohen
Regression example
• Generated: w=2 • Recovered: w=2.08 • Noise: std=4
Slide courtesy of William Cohen
Bias term • So far we assumed that the
line passes through the origin • What if the line does not? • No problem, simply change the
model to y = w0 + w1x+ε • Can use least squares to
determine w0 , w1
n
xwyw i
ii∑ −=
1
0
X
Y
w0
∑
∑ −=
ii
iii
x
wyxw 2
0
1
)(
Slide courtesy of William Cohen
Linear Regression
12
Data: Inputs are continuous vectors of length K. Outputs are continuous scalars.
D = {x(i), y(i)}Ni=1 where x � RK and y � R
What are some example problems of
this form?
Linear Regression
13
Data: Inputs are continuous vectors of length K. Outputs are continuous scalars.
D = {x(i), y(i)}Ni=1 where x � RK and y � R
Prediction: Output is a linear function of the inputs. y = h�(x) = �1x1 + �2x2 + . . . + �KxK
y = h�(x) = �T x (We assume x1 is 1)
Learning: finds the parameters that minimize some objective function.
�� = argmin�
J(�)
Least Squares
14
Learning: finds the parameters that minimize some objective function. We minimize the sum of the squares: Why?
1. Reduces distance between true measurements and predicted hyperplane (line in 1D)
2. Has a nice probabilistic interpretation
�� = argmin�
J(�)
J(�) =1
2
N�
i=1
(�T x(i) � y(i))2
Least Squares
15
Learning: finds the parameters that minimize some objective function. We minimize the sum of the squares: Why?
1. Reduces distance between true measurements and predicted hyperplane (line in 1D)
2. Has a nice probabilistic interpretation
�� = argmin�
J(�)
J(�) =1
2
N�
i=1
(�T x(i) � y(i))2This is a very general optimization setup. We could solve it in lots of ways. Today, we’ll consider three
ways.
Least Squares
16
Learning: Three approaches to solving
Approach 2: Stochastic Gradient Descent (SGD) (take many small steps opposite the gradient)
Approach 1: Gradient Descent (take larger – more certain – steps opposite the gradient)
Approach 3: Closed Form (set derivatives equal to zero and solve for parameters)
�� = argmin�
J(�)
Learning as optimization
• The main idea today – Make two assumptions:
• the classifier is linear, like naïve Bayes – ie roughly of the form fw(x) = sign(x . w)
• we want to find the classifier w that “fits the data best”
– Formalize as an optimization problem • Pick a loss function J(w,D), and find
argminw J(w) • OR: Pick a “goodness” function, often Pr(D|w), and find the argmaxw fD(w)
17 Slide courtesy of William Cohen
Learning as optimization: warmup Find: optimal (MLE) parameter θ of a binomial Dataset: D={x1,…,xn}, xi is 0 or 1, k of them are 1
18
P(D |θ ) = const θ xi (1−i∏ θ )1−xi =θ k (1−θ )n−k
ddθ
P(D |θ ) = ddθ
θ k (1−θ )n−k( )
=ddθθ k"
#$
%
&'(1−θ )n−k +θ k d
dθ(1−θ )n−k
= kθ k−1(1−θ )n−k +θ k (n− k)(1−θ )n−k−1(−1)
=θ k−1(1−θ )n−k−1 k(1−θ )−θ(n− k)( )Slide courtesy of William Cohen
Learning as optimization: warmup Goal: Find the best parameter θ of a binomial Dataset: D={x1,…,xn}, xi is 0 or 1, k of them are 1
19
= 0
θ= 0 θ= 1 k-‐ kθ – nθ + kθ = 0
è nθ = k è θ = k/n
Slide courtesy of William Cohen
Learning as optimization with gradient ascent
• Goal: Learn the parameter w of …
• Dataset: D={(x1,y1)…,(xn , yn)} • Use your model to define
– Pr(D|w) = …. • Set w to maximize Likelihood
– Usually we use numeric methods to find the optimum
– i.e., gradient ascent: repeatedly take a small step in the direction of the gradient
20
Difference between ascent and descent is only the sign: I’m going to be sloppy here about that.
Slide courtesy of William Cohen
Gradient ascent
21
To find argminx f(x): • Start with x0 • For t=1….
• xt+1 = xt + λ f’(x t) where λ is small
Slide courtesy of William Cohen
Gradient descent
22
Likelihood: ascent Loss: descent
Slide courtesy of William Cohen
Pros and cons of gradient descent
• Simple and often quite effective on ML tasks • Often very scalable • Only applies to smooth functions (differentiable) • Might find a local minimum, rather than a global
one
23 Slide courtesy of William Cohen
Pros and cons of gradient descent
24
There is only one local optimum if the function is convex
Slide courtesy of William Cohen
Gradient Descent
25
Algorithm 1 Gradient Descent
1: procedure GD(D, �(0))2: � � �(0)
3: while not converged do4: � � � + ���J(�)
5: return �
In order to apply GD to Linear Regression all we need is the gradient of the objective function (i.e. vector of partial derivatives).
��J(�) =
�
����
dd�1
J(�)d
d�2J(�)...
dd�N
J(�)
�
����
Gradient Descent
26
Algorithm 1 Gradient Descent
1: procedure GD(D, �(0))2: � � �(0)
3: while not converged do4: � � � + ���J(�)
5: return �
There are many possible ways to detect convergence. For example, we could check whether the L2 norm of the gradient is below some small tolerance.
||��J(�)||2 � �Alternatively we could check that the reduction in the objective function from one iteration to the next is small.
Stochastic Gradient Descent (SGD)
27
Algorithm 2 Stochastic Gradient Descent (SGD)
1: procedure SGD(D, �(0))2: � � �(0)
3: while not converged do4: for i � shu�e({1, 2, . . . , N}) do5: � � � + ���J (i)(�)
6: return �
We need a per-‐example objective:
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
Applied to Linear Regression, SGD is called the Least Mean Squares (LMS) algorithm
Stochastic Gradient Descent (SGD)
We need a per-‐example objective:
28
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
Applied to Linear Regression, SGD is called the Least Mean Squares (LMS) algorithm
Algorithm 2 Stochastic Gradient Descent (SGD)
1: procedure SGD(D, �(0))2: � � �(0)
3: while not converged do4: for i � shu�e({1, 2, . . . , N}) do5: for k � {1, 2, . . . , K} do6: �k � �k + � d
d�kJ (i)(�)
7: return �
Stochastic Gradient Descent (SGD)
We need a per-‐example objective:
29
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
Applied to Linear Regression, SGD is called the Least Mean Squares (LMS) algorithm
Algorithm 2 Stochastic Gradient Descent (SGD)
1: procedure SGD(D, �(0))2: � � �(0)
3: while not converged do4: for i � shu�e({1, 2, . . . , N}) do5: for k � {1, 2, . . . , K} do6: �k � �k + � d
d�kJ (i)(�)
7: return �
Let’s start by calculating this partial derivative for the Linear Regression objective function.
Partial Derivatives for Linear Reg.
30
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
Partial Derivatives for Linear Reg.
31
d
d�kJ (i)(�) =
d
d�k
1
2(�T x(i) � y(i))2
=1
2
d
d�k(�T x(i) � y(i))2
= (�T x(i) � y(i))d
d�k(�T x(i) � y(i))
= (�T x(i) � y(i))d
d�k(
K�
k=1
�kx(i)k � y(i))
= (�T x(i) � y(i))x(i)k
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
Partial Derivatives for Linear Reg.
32
Let J(�) =�N
i=1 J (i)(�)where J (i)(�) = 1
2 (�T x(i) � y(i))2.
d
d�kJ (i)(�) = (�T x(i) � y(i))x(i)
k
d
d�kJ(�) =
d
d�k
N�
i=1
J (i)(�)
=N�
i=1
(�T x(i) � y(i))x(i)k
Used by SGD (aka. LMS)
Used by Gradient Descent
Least Mean Squares (LMS)
33
Applied to Linear Regression, SGD is called the Least Mean Squares (LMS) algorithm
Algorithm 3 Least Mean Squares (LMS)
1: procedure LMS(D, �(0))2: � � �(0)
3: while not converged do4: for i � shu�e({1, 2, . . . , N}) do5: for k � {1, 2, . . . , K} do
6: �k � �k + �(�T x(i) � y(i))x(i)k
7: return �
Least Squares
34
Learning: Three approaches to solving
Approach 2: Stochastic Gradient Descent (SGD) (take many small steps opposite the gradient)
Approach 1: Gradient Descent (take larger – more certain – steps opposite the gradient)
Approach 3: Closed Form (set derivatives equal to zero and solve for parameters)
�� = argmin�
J(�)
Background: Matrix Derivatives l For , define:
l Trace:
l Some fact of matrix derivatives (without proof)
© Eric Xing @ CMU, 2006-2011 35
⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
∂
∂
∂
∂
∂
∂
∂
∂
=∇
fA
fA
fA
fA
Af
mnm
n
A
!
"#"
!
1
111
)(
RR !nmf ×:
, tr ∑=
=n
iiiAA
1, tr aa = BCACABABC trtrtr ==
, tr TA BAB =∇ , tr TTT
A ABCCABCABA +=∇ ( ) TA AAA 1−=∇
The normal equations l Write the cost function in matrix form:
l To minimize J(θ), take derivative and set to zero:
© Eric Xing @ CMU, 2006-2011 36
( ) ( )
( )yyXyyXXX
yXyX
yJ
TTTTTT
T
n
ii
Ti
!!!!
!!
+−−=
−−=
−= ∑=
θθθθ
θθ
θθ
212121
1
2)()( x
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−−−−
−−−−
−−−−
=
nx
xx
X!!!
2
1
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
=
ny
yy
y!
" 2
1
( )
( )
( )0
221
22121
=−=
−+=
∇+∇−∇=
+−−∇=∇
yXXX
yXXXXX
yyXyXX
yyXyyXXXJ
TT
TTT
TTTT
TTTTTT
!
!
!!!
!!!!
θ
θθ
θθθ
θθθθ
θθθ
θθ
trtrtr
tryXXX TT !
=⇒ θ The normal equations
( ) yXXX TT !1−=*θ
⇓
Comments on the normal equation
l In most situations of practical interest, the number of data points N is larger than the dimensionality k of the input space and the matrix X is of full column rank. If this condition holds, then it is easy to verify that XTX is necessarily invertible.
l The assumption that XTX is invertible implies that it is positive definite, thus the critical point we have found is a minimum.
l What if X has less than full column rank? à regularization (later).
© Eric Xing @ CMU, 2006-2011 37
Direct and Iterative methods l Direct methods: we can achieve the solution in a single step
by solving the normal equation l Using Gaussian elimination or QR decomposition, we converge in a finite number
of steps l It can be infeasible when data are streaming in in real time, or of very large
amount
l Iterative methods: stochastic or steepest gradient l Converging in a limiting sense l But more attractive in large practical problems l Caution is needed for deciding the learning rate α
© Eric Xing @ CMU, 2006-2011 38
Convergence rate l Theorem: the steepest descent equation algorithm
converge to the minimum of the cost characterized by normal equation:
If
l A formal analysis of LMS needs more math; in practice,
one can use a small α, or gradually decrease α.
© Eric Xing @ CMU, 2006-2011 39
Convergence Curves
l For the batch method, the training MSE is initially large due to uninformed initialization
l In the online update, N updates for every epoch reduces MSE to a much smaller value.
40 © Eric Xing @ CMU, 2006-2011
Least Squares
41
Learning: Three approaches to solving
Approach 2: Stochastic Gradient Descent (SGD) (take many small steps opposite the gradient)
• pros: memory efficient, fast convergence, less prone to local optima
• cons: convergence in practice requires tuning and fancier variants
Approach 1: Gradient Descent (take larger – more certain – steps opposite the gradient)
• pros: conceptually simple, guaranteed convergence • cons: batch, often slow to converge
Approach 3: Closed Form (set derivatives equal to zero and solve for parameters)
• pros: one shot algorithm! • cons: does not scale to large datasets (matrix inverse is
bottleneck)
�� = argmin�
J(�)
Geometric Interpretation of LMS • The predictions on the training data are:
• Note that
and is the orthogonal projection of
into the space spanned by the columns of X
( ) yXXXXXy TT !! 1−== *ˆ θ
( )( )yIXXXXyy TT !!!−=−
−1ˆ
( ) ( )( )( )( )
0
1
1
=
−=
−=−−
−
yXXXXXX
yIXXXXXyyXTTTT
TTTT
!
!!!
!!
y! y!
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−−−−
−−−−
−−−−
=
nx
xx
X!!!
2
1
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
=
ny
yy
y!
" 2
1
42 © Eric Xing @ CMU, 2006-‐2011
Probabilistic Interpretation of LMS • Let us assume that the target variable and the inputs
are related by the equation:
where ε is an error term of unmodeled effects or random
noise
• Now assume that ε follows a Gaussian N(0,σ), then we have:
• By independence assumption:
iiT
iy εθ += x
⎟⎟⎠
⎞⎜⎜⎝
⎛ −−= 2
2
221
σθ
σπθ
)(exp);|( iT
iii
yxyp x
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛ −−⎟
⎠
⎞⎜⎝
⎛==
∑∏ =
=2
12
1 221
σ
θ
σπθθ
n
i iT
inn
iii
yxypL
)(exp);|()(
x
43 © Eric Xing @ CMU, 2006-‐2011
Probabilistic Interpretation of LMS • Hence the log-‐likelihood is:
• Do you recognize the last term? Yes it is:
• Thus under independence assumption, LMS is equivalent to MLE of θ !
∑ =−−=
n
i iT
iynl1
22 211
21 )(log)( xθ
σσπθ
∑=
−=n
ii
Ti yJ
1
2
21 )()( θθ x
44 © Eric Xing @ CMU, 2006-‐2011
Non-Linear basis function • So far we only used the observed values x1,x2,… • However, linear regression can be applied in the same
way to functions of these values – Eg: to add a term w x1x2 add a new variable z=x1x2 so each
example becomes: x1, x2, …. z
• As long as these functions can be directly computed from the observed values the parameters are still linear in the data and the problem remains a multi-variate linear regression problem
ε++++= 22110 kk xwxwwy …
Slide courtesy of William Cohen
Non-linear basis functions
• What type of functions can we use? • A few common examples: - Polynomial: φj(x) = xj for j=0 … n - Gaussian: - Sigmoid:
- Logs:
€
φ j (x) =(x −µ j )2σ j
2
€
φ j (x) =1
1+ exp(−s j x)
Any function of the input values can be used. The solution for the parameters of the regression remains the same.
φ j (x) = log(x +1)
Slide courtesy of William Cohen
General linear regression problem • Using our new notations for the basis function linear
regression can be written as
• Where φj(x) can be either xj for multivariate regression or one of the non-linear basis functions we defined
• … and φ0(x)=1 for the intercept term €
y = w jφ j (x)j= 0
n
∑
Slide courtesy of William Cohen
An example: polynomial basis vectors on a small dataset
– From Bishop Ch 1
Slide courtesy of William Cohen
0th Order Polynomial
n=10 Slide courtesy of William Cohen
1st Order Polynomial
Slide courtesy of William Cohen
3rd Order Polynomial
Slide courtesy of William Cohen
9th Order Polynomial
Slide courtesy of William Cohen
Over-fitting
Root-Mean-Square (RMS) Error:
Slide courtesy of William Cohen
Polynomial Coefficients
Slide courtesy of William Cohen
Regularization
Penalize large coefficient values
JX,y (w) =12
yi − wjφ j (xi )
j∑
#
$%%
&
'((
i∑
2
−λ2w 2
Slide courtesy of William Cohen
Regularization: +
Slide courtesy of William Cohen
Polynomial Coefficients
none exp(18) huge
Slide courtesy of William Cohen
Over Regularization:
Slide courtesy of William Cohen
Example: Stock Prices • Suppose we wish to predict Google’s stock price at time t+1
• What features should we use? (putting all computational concerns aside) – Stock prices of all other stocks at times t, t-‐1, t-‐2, …, t -‐ k
– Mentions of Google with positive / negative sentiment words in all newspapers and social media outlets
• Do we believe that all of these features are going to be useful?
59
Ridge Regression
• Adds an L2 regularizer to Linear Regression
• Bayesian interpretation: MAP estimation with a Gaussian prior on the parameters
60
JRR(�) = J(�) + �||�||22
=1
2
N�
i=1
(�T x(i) � y(i))2 + �K�
k=1
�2k
�MAP = argmax�
N�
i=1
log p�(y(i)|x(i)) + log p(�)
= argmax�
JRR(�)where
p(�) � N (0,1�)
�MAP = argmax�
N�
i=1
log p�(y(i)|x(i)) + log p(�)
= argmax�
JLASSO(�)
JLASSO(�) = J(�) + �||�||1
=1
2
N�
i=1
(�T x(i) � y(i))2 + �K�
k=1
|�k|
LASSO
• Adds an L1 regularizer to Linear Regression
• Bayesian interpretation: MAP estimation with a Laplace prior on the parameters
61
where
p(�) � Laplace(0, f(�))
Ridge Regression vs Lasso
Ridge Regression: Lasso:
Lasso (l1 penalty) results in sparse solu3ons – vector with more zero coordinates Good for high-‐dimensional problems – don’t have to store all coordinates!
βs with constant l1 norm
βs with constant J(β) (level sets of J(β))
βs with constant l2 norm
β2
β1
62 © Eric Xing @ CMU, 2006-‐2011
X X
Optimization for LASSO
Can we apply SGD to the LASSO learning problem?
63
JLASSO(�) = J(�) + �||�||1
=1
2
N�
i=1
(�T x(i) � y(i))2 + �K�
k=1
|�k|
�MAP = argmax�
N�
i=1
log p�(y(i)|x(i)) + log p(�)
= argmax�
JLASSO(�)
Optimization for LASSO
• Consider the absolute value function:
64
r(�) = �K�
k=1
|�k|
• Where does a small step opposite the gradient accomplish… – … far from zero? – … close to zero?
• The L1 penalty is subdifferentiable (i.e. not differentiable at 0)
Optimization for LASSO • The L1 penalty is subdifferentiable (i.e. not differentiable at 0)
• An array of optimization algorithms exist to handle this issue: – Coordinate Descent – Othant-‐Wise Limited memory Quasi-‐Newton (OWL-‐QN) (Andrew & Gao, 2007) and provably convergent variants
– Block coordinate Descent (Tseng & Yun, 2009) – Sparse Reconstruction by Separable Approximation (SpaRSA) (Wright et al., 2009)
– Fast Iterative Shrinkage Thresholding Algorithm (FISTA) (Beck & Teboulle, 2009)
65
Data Set Size: 9th Order Polynomial
Slide courtesy of William Cohen
Locally weighted linear regression
l The algorithm: Instead of minimizing now we fit θ to minimize Where do wi's come from?
l where x is the query point for which we'd like to know its corresponding y
à Essentially we put higher weights on (errors on) training examples that are close to the query point (than those that are further away from the query)
© Eric Xing @ CMU, 2006-2011 67
∑=
−=n
ii
Ti yJ
1
2
21 )()( θθ x
∑=
−=n
ii
Tii ywJ
1
2
21 )()( θθ x
⎟⎟⎠
⎞⎜⎜⎝
⎛ −−= 2
2
2τ)(exp xxi
iw
Parametric vs. non-parametric l Locally weighted linear regression is the second example we are
running into of a non-parametric algorithm. (what is the first?)
l The (unweighted) linear regression algorithm that we saw earlier is known as a parametric learning algorithm l because it has a fixed, finite number of parameters (the θ), which are fit to the data; l Once we've fit the θ and stored them away, we no longer need to keep the training
data around to make future predictions. l In contrast, to make predictions using locally weighted linear regression, we need to
keep the entire training set around.
l The term "non-parametric" (roughly) refers to the fact that the amount of stuff we need to keep in order to represent the hypothesis grows linearly with the size of the training set.
© Eric Xing @ CMU, 2006-2011 68
Robust Regression
l The best fit from a quadratic regression
l But this is probably better …
© Eric Xing @ CMU, 2006-2011 69
How can we do this?
LOESS-based Robust Regression l Remember what we do in "locally weighted linear regression"?
à we "score" each point for its impotence l Now we score each point according to its "fitness"
© Eric Xing @ CMU, 2006-2011 70
(Courtesy to Andrew Moore)
Robust regression l For k = 1 to R…
l Let (xk ,yk) be the kth datapoint l Let yest
k be predicted value of yk
l Let wk be a weight for data point k that is large if the data point fits well and small if it fits badly:
l Then redo the regression using weighted data points.
l Repeat whole thing until converged!
© Eric Xing @ CMU, 2006-2011 71
( )2)( estkkk yyw −= φ
Robust regression—probabilistic interpretation
l What regular regression does:
Assume yk was originally generated using the following recipe:
Computational task is to find the Maximum Likelihood estimation of θ
© Eric Xing @ CMU, 2006-2011 72
),( 20 σθ N+= kT
ky x
Robust regression—probabilistic interpretation
l What LOESS robust regression does:
Assume yk was originally generated using the following recipe:
with probability p:
but otherwise
Computational task is to find the Maximum Likelihood estimates of θ, p, µ and σhuge.
l The algorithm you saw with iterative reweighting/refitting does this computation for us. Later you will find that it is an instance of the famous E.M. algorithm
© Eric Xing @ CMU, 2006-2011 73
),( 20 σθ N+= kT
ky x
),(~ huge2σµNky
Summary
• Linear regression predicts its output as a linear function of its inputs
• Learning optimizes a function (equivalently likelihood or mean squared error) using standard techniques (gradient descent, SGD, closed form)
• Regularization enables shrinkage and model selection
© Eric Xing @ CMU, 2005-‐2015 74