Bandwidth selection for kernel conditional density estimation

By David M. Bashtannyk, Rob J Hyndman. 2001

298 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Gerhard E. Tutz
2. Jochen Einbeck
3. Jeffrey S. Racine
4. Qi Li
5. Sam Efromovich
6. Richard E. Turner
7. Qiwei Yao
8. Brian L. Trippe

In Control and Systems Engineering

1. Jeffrey S. Racine
2. Juan Lin
3. Qi Li
4. Jochen Einbeck
5. Qiwei Yao
6. Gerhard E. Tutz
7. Rob J Hyndman
8. Sam Efromovich

In Statistics and Probability

1. Jeffrey S. Racine
2. Qi Li
3. Gerhard E. Tutz
4. Jochen Einbeck
5. Juan Lin
6. Sam Efromovich
7. Qiwei Yao
8. Rob J Hyndman

Citations in the corpus

• CONDITIONAL DENSITY ESTIMATION IN A REGRESSION SETTING (2007, background): “The interested reader can find a historical overview of this and related approaches in the books [5, 15, 34]; other relevant references are [2, 16, 17, 21, 22, 23, 26, 27, 28, 35].”
• Conditional Density Estimation with Bayesian Normalising Flows (2018, background): “In particular, nonparametric meth-ods require passing through the entire dataset for each prediction and demand one to set potentially arbitrary bounds to filtering points to be included in calculation of conditional densities (Bashtannyk & Hyndman, 2001).”
• Optimal Bandwidth Selection for Nonparametric Conditional Distribution and Quantile Functions (2013, methodology): “(1976), ‘Multivariate binary discrimination by the kernel method’, Biometrika 63(3), 413– 420. Bashtannyk, D. M. & Hyndman, R. J. (2001), ‘Bandwidth selection for kernel conditional density estimation’, Computational Statistics and Data Analysis 36, 279–298. Cai, Z. (2002), ‘Regression quantiles for”
• Modelling beyond regression functions: an application of multimodal regression to speed–flow data (2006, methodology): “Looking at box plots of high density regions (Figs 8(a) and 8(b)), we note that these bandwidths are obviously oversmoothing, as was also observed by Bashtannyk and Hyndman (2001) and Fan and Yim (2004).”
• Nonparametric Estimation and Symmetry Tests for Conditional Density Functions (2002, methodology): “…f 2ðxÞ dx dy ¼ t 2 r nhb þ arh2r1 þ brhr1 b2 þ gb4 1 þ oð1Þ ð3:1Þ where ar ¼ Z2r ð ð gðjr1ÞðyjxÞ Aðr1Þ 0; yxy 2 f 2ðxÞ dx dy ð3:2Þ br ¼ m2Zr ð ð gð2ÞðyjxÞ gðjr1ÞðyjxÞ Aðr1Þ 0; yxy f 2ðxÞ dx dy ð3:3Þ and g ¼ m 2 2 4 ð ð gð2ÞðyjxÞ 2 f 2ðxÞ dx dy: ð3:4Þ Bashtannyk and Hyndman (2001) used a similar…”