Upload
others
View
11
Download
0
Embed Size (px)
Citation preview
EE613Machine Learning for Engineers
NONLINEAR REGRESSION I
Sylvain CalinonRobot Learning & Interaction Group
Idiap Research InstituteDec. 13, 2017
1
Outline
2
• Properties of multivariate Gaussian distributions Matlab codes: demo_Gaussian_product01.m, demo_Gaussian_conditioning01.m, demo_Gaussian_conditioning_noisyInput01.m, demo_Gaussian_lawTotalCov01.m
• Locally weighted regression (LWR)Matlab code: demo_LWR01.m,
• Dynamical movement primitives (DMP)Matlab code: demo_DMP01.m
• Gaussian mixture regression (GMR)Matlab codes: demo_GMR01.m, demo_GMR_polyFit01.m, demo_DMP_GMR01.m
Properties of multivariate Gaussian distributions
Matlab codes: demo_Gaussian_product01.m,
demo_Gaussian_conditioning01.m,
demo_Gaussian_conditioning_noisyInput01.m,
demo_Gaussian_lawTotalCov01.m
3
The Matrix Cookbook
Kaare Brandt Petersen
Michael SyskindPedersen
Product of Gaussians:
Linear combination:
Conditional probability:
Some very useful properties
4
Product of Gaussians
Linear combination
6
in a new situation…
Example: Fusion of sensor/control information
Coordinate system 1:This is where I expect
data to be located!
Coordinate system 2:This is where I expect
data to be located!
Product of linearly transformed Gaussians
[Calinon, S. (2016). A Tutorial on Task-Parameterized Movement Learning and Retrieval. Intelligent Service Robotics 9(1):1–29]
7
Conditional probability
8→ Linear regression from joint distribution
9
Conditional probability
Conditional probability - Geometric interpretation
10
11
Conditional probability - Proof
12
Conditional probability - Proof
13
Conditional probability - Proof
14
Conditional probability - Proof
15
Conditional probability - Proof
16
Conditional probability - Summary
17
Gaussian conditioning with uncertain inputs
18
Gaussian conditioning with uncertain inputs
Gaussian estimate of a mixture of Gaussians
19
20
Gaussian estimate of a mixture - Proof
Gaussian estimate of a mixture of Gaussians
21
Locally weighted regression(LWR)
Matlab codes: demo_LWR01.m, demo_DMP01.m
22
[C.G. Atkeson, A.W. Moore, and S. Schaal. Locally weighted learning for control. Artificial Intelligence Review, 11(1-5):75–113, 1997]
[W.S. Cleveland. Robust locally weighted regression and smoothing scatterplots. American Statistical Association 74(368):829–836, 1979]
Previous lecture on linear regression
23
Color darknessproportional
to weight
Locally weighted regression (LWR)
24
Locally weighted regression (LWR)
25
Locally weighted regression (LWR)
26
LWR can be used for local least squares polynomial
fitting by changing the definition of the inputs.
Locally weighted regression (LWR)
27
Locally weighted regression (LWR)
28
Dynamical movement primitives (DMP)
¿ _s = ¡®s
[Ijspeert, Nakanishi and Schaal, NIPS’2003]
[Ijspeert, Nakanishi, Pastor, Hoffmann and Schaal, Neural Computation 25(2), 2013]
Set of basis functionsDecay termPosition Velocity Acceleration
Locally weighted
regression
(LWR)
INP
UT
SO
UT
PU
TS
Gaussian mixture regression(GMR)
Matlab codes: demo_GMR01.mdemo_GMR_polyFit01.m, demo_DMP_GMR01.m
30
[Z. Ghahramani and M. I. Jordan. Supervised learning from incomplete data via an EM approach. In Advances in Neural Information Processing Systems (NIPS), volume 6, pages 120–127, 1994]
[S. Calinon S. A tutorial on task-parameterized movement learning and retrieval. Intelligent Service Robotics 9(1):1–29, 2016]
Gaussian mixture regression (GMR)
31
32
Gaussian mixture regression (GMR)
• Gaussian mixture regression (GMR) is a nonlinear regression technique that does not model the regression function directly, but instead first models the joint probability density of input-output data in the form of a Gaussian mixture model (GMM).
• The computation relies on linear transformation and conditioning properties of multivariate normal distributions.
• GMR provides a regression approach in which multivariate output distributions can be computed in an online manner, with a computation time independent of the number of datapoints used to train the model, by exploiting the learned joint density model.
• In GMR, both input and output variables can be multivariate, and after learning, any subset of input-output dimensions can be selected for regression. This can for example be exploited to handle different sources of missing data, where expectations on the remaining dimensions can be computed as a multivariate distribution.
33
Gaussian mixture regression (GMR)
34
Gaussian mixture regression (GMR)
35
Gaussian mixture regression (GMR) - Proof
GMR can cover a large spectrum of regression mechanisms
Both and can be multidimensional
encoded in Gaussian mixture model (GMM)
retrieved by Gaussian mixture regression (GMR)
Gaussian mixture regression (GMR)
Nadaraya-Watsonkernel regression
Least squareslinear regression
36
37
GMR with uncertain inputs
GMR for smooth piecewise polynomial fitting
38
[Calinon, Guenter and Billard,
IEEE Trans. on SMC-B 37(2), 2007]
Gaussian mixture regression - Examples
[Hersch, Guenter, Calinon and Billard,
IEEE Trans. on Robotics 24(6), 2008]
With expectation-maximization (EM): (maximizing log-likelihood)
[Khansari-Zadeh and Billard,
IEEE Trans. on Robotics 27(5), 2011]
With quadratic programming solver:(maximizing log-likelihood s.t. stability constraints)
Learning of and retrieval of
Dynamical movement primitives with GMR
40
Main references
Regression
F. Stulp and O. Sigaud. Many regression algorithms, one unified model – a review. Neural Networks, 69:60–79, 2015
LWR
C. G. Atkeson, A. W. Moore, and S. Schaal. Locally weighted learning for control. Artificial Intelligence Review, 11(1-5):75–113, 1997
W.S. Cleveland. Robust locally weighted regression and smoothing scatterplots. American Statistical Association 74(368):829–836, 1979
Recursive formulation of LWR
S. Schaal and C.G. Atkeson. Constructive incremental learning from only local information. Neural Computation 10(8):2047–2084, 1998.
Bayesian formulation of LWR
J. Ting, M. Kalakrishnan, S. Vijayakumar and S. Schaal. Bayesian kernel shaping for learning control. In: Advances in Neural Information Processing Systems (NIPS), pp 1673–1680, 2008.
41
Main references
DMPA. Ijspeert, J. Nakanishi, P. Pastor, H. Hoffmann, and S. Schaal. Dynamical movement primitives: Learning attractor models for motor behaviors. Neural Computation, 25(2):328–373, 2013
LWPR
S. Vijayakumar, A. D’souza and S. Schaal. Incremental online learning in high dimensions. Neural Computation 17(12):2602–2634, 2005
GMR
Z. Ghahramani and M. I. Jordan. Supervised learning from incomplete data via an EM approach. In Advances in Neural Information Processing Systems (NIPS), volume 6, pages 120–127, 1994
S. Calinon S. A tutorial on task-parameterized movement learning and retrieval. Intelligent Service Robotics 9(1):1–29, 2016
42