Uncertainty in Artificial Intelligence
First Name   Last Name   Password   Forgot Password   Log in!
    Proceedings         Authors   Author's Info   Article details         Search    
Faster Gaussian Summation: Theory and Experiment
Dongryeol Lee, Alexander Gray
Abstract:
We provide faster algorithms for the problem of Gaussian summation, which occurs in many machine learning methods. We develop two new extensions - an O(Dp) Taylor expansion for the Gaussian kernel with rigorous error bounds and a new error control scheme integrating any arbitrary approximation method - within the best discretealgorithmic framework using adaptive hierarchical data structures. We rigorously evaluate these techniques empirically in the context of optimal bandwidth selection in kernel density estimation, revealing the strengths and weaknesses of current state-of-the-art approaches for the first time. Our results demonstrate that the new error control scheme yields improved performance, whereas the series expansion approach is only effective in low dimensions (five or less).
Keywords:
Pages: 281-288
PS Link:
PDF Link: /papers/06/p281-lee.pdf
BibTex:
@INPROCEEDINGS{Lee06,
AUTHOR = "Dongryeol Lee and Alexander Gray",
TITLE = "Faster Gaussian Summation: Theory and Experiment",
BOOKTITLE = "Proceedings of the Twenty-Second Conference Annual Conference on Uncertainty in Artificial Intelligence (UAI-06)",
PUBLISHER = "AUAI Press",
ADDRESS = "Arlington, Virginia",
YEAR = "2006",
PAGES = "281--288"
}


hosted by DSL   •   site info   •   help