K-mappings and Regression trees

Yi Wang; Arthur Szlam

doi:10.1109/ICASSP.2014.6854138

K-mappings and Regression trees

Yi Wang, Arthur Szlam

Research output: Chapter in Book/Entry/Poem › Conference contribution

Abstract

We describe a method for learning a piecewise affine approximation to a mapping f : R^d → R^p given a labeled training set of examples {x₁,..., x_n} = X ⊂ R^d and targets {y₁ = f(x₁),..., y_n = f(x_n)} = Y ⊂ R^p. The method first trains a binary subdivision tree that splits across hyperplanes in X corresponding to high variance directions in Y. A fixed number K of affine regressors of rank q are then trained via a K-means like iterative algorithm, where each leaf must vote on its best fit mapping, and each mapping is updated as the best fit for the collection of leaves that chose it.

Original language	English (US)
Title of host publication	2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	2937-2941
Number of pages	5
ISBN (Print)	9781479928927
DOIs	https://doi.org/10.1109/ICASSP.2014.6854138
State	Published - 2014
Externally published	Yes
Event	2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014 - Florence, Italy Duration: May 4 2014 → May 9 2014

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN (Print)	1520-6149

Other

Other	2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014
Country/Territory	Italy
City	Florence
Period	5/4/14 → 5/9/14

Keywords

Partial Least squares
Piecewise linear Regression
Sparse Modeling

ASJC Scopus subject areas

Software
Signal Processing
Electrical and Electronic Engineering

Access to Document

10.1109/ICASSP.2014.6854138

Cite this

Wang, Y., & Szlam, A. (2014). K-mappings and Regression trees. In 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014 (pp. 2937-2941). Article 6854138 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP.2014.6854138

K-mappings and Regression trees. / Wang, Yi; Szlam, Arthur.
2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 2937-2941 6854138 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Wang, Y & Szlam, A 2014, K-mappings and Regression trees. in 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014., 6854138, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 2937-2941, 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014, Florence, Italy, 5/4/14. https://doi.org/10.1109/ICASSP.2014.6854138

@inproceedings{bed8536f69754ae0955f7a5d097e4bf4,

title = "K-mappings and Regression trees",

abstract = "We describe a method for learning a piecewise affine approximation to a mapping f : Rd → Rp given a labeled training set of examples {x1,..., xn} = X ⊂ Rd and targets {y1 = f(x1),..., yn = f(xn)} = Y ⊂ Rp. The method first trains a binary subdivision tree that splits across hyperplanes in X corresponding to high variance directions in Y. A fixed number K of affine regressors of rank q are then trained via a K-means like iterative algorithm, where each leaf must vote on its best fit mapping, and each mapping is updated as the best fit for the collection of leaves that chose it.",

keywords = "Partial Least squares, Piecewise linear Regression, Sparse Modeling",

author = "Yi Wang and Arthur Szlam",

year = "2014",

doi = "10.1109/ICASSP.2014.6854138",

language = "English (US)",

isbn = "9781479928927",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "2937--2941",

booktitle = "2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014",

note = "2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014 ; Conference date: 04-05-2014 Through 09-05-2014",

}

TY - GEN

T1 - K-mappings and Regression trees

AU - Wang, Yi

AU - Szlam, Arthur

PY - 2014

Y1 - 2014

N2 - We describe a method for learning a piecewise affine approximation to a mapping f : Rd → Rp given a labeled training set of examples {x1,..., xn} = X ⊂ Rd and targets {y1 = f(x1),..., yn = f(xn)} = Y ⊂ Rp. The method first trains a binary subdivision tree that splits across hyperplanes in X corresponding to high variance directions in Y. A fixed number K of affine regressors of rank q are then trained via a K-means like iterative algorithm, where each leaf must vote on its best fit mapping, and each mapping is updated as the best fit for the collection of leaves that chose it.

AB - We describe a method for learning a piecewise affine approximation to a mapping f : Rd → Rp given a labeled training set of examples {x1,..., xn} = X ⊂ Rd and targets {y1 = f(x1),..., yn = f(xn)} = Y ⊂ Rp. The method first trains a binary subdivision tree that splits across hyperplanes in X corresponding to high variance directions in Y. A fixed number K of affine regressors of rank q are then trained via a K-means like iterative algorithm, where each leaf must vote on its best fit mapping, and each mapping is updated as the best fit for the collection of leaves that chose it.

KW - Partial Least squares

KW - Piecewise linear Regression

KW - Sparse Modeling

UR - http://www.scopus.com/inward/record.url?scp=84905260056&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84905260056&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2014.6854138

DO - 10.1109/ICASSP.2014.6854138

M3 - Conference contribution

AN - SCOPUS:84905260056

SN - 9781479928927

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 2937

EP - 2941

BT - 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014

Y2 - 4 May 2014 through 9 May 2014

ER -

K-mappings and Regression trees

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this