Abstract
This paper describes “bumptrees”, a new approach to improving the computational efficiency of a wide variety of connectionist algorithms. We describe the use of these structures for representing, learning, and evaluating smooth mappings, smooth constraints, classification regions, and probability densities. We present an empirical comparison of a bumptree approach to more traditional connectionist approaches for learning the mapping between the kinematic and visual representations of the state of a 3 joint robot arm. Simple networks based on backpropagation with sigmoidal units are unable to perform the task at all. Radial basis function networks perform the task but by using bumptrees, the learning rate is hundreds of times faster at reasonable error levels and the retrieval time is over fifty times faster with 10,000 samples. Bumptrees are a natural generalization of oct-trees, k-d trees, balltrees and boxtrees and are useful in a variety of circumstances. We describe both the underlying ideas and extensions to constraint and classification learning that are under current investigation.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
L. Devroye and L. Gyorfi. (1985) Nonparametric Density Estimation: The LI View, New York: Wiley.
J. H. Friedman, J. L. Bendey and R. A. Finkel. (1977) An algorithm for finding best matches in logarithmic expected time. ACM Trans. Math. Software 3: 209–226.
B. Mel. (1990) Connectionist Robot Motion Planning, A Neurally-Inspired Approach to Visually-Guided Reaching, San Diego, CA: Academic Press.
S. M. Omohundro. (1987) Efficient algorithms with neural network behavior. Complex Systems 1: 273–347.
S. M. Omohundro. (1989) Five balltree construction algorithms. International Computer Science Institute Technical Report TR-89-063.
S. M. Omohundro. (1990) Geometric learning algorithms. Physica D 42: 307–321.
R. F. Sproull. (1990) Refinements to Nearest-Neighbor Searching in k-d Trees. Sutherland, Sproull and Associates Technical Report SSAPP #184c, to appear in Algorithmica.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1991 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Omohundro, S.M. (1991). Building Faster Connectionist Systems With Bumptrees. In: Brauer, W., Hernández, D. (eds) Verteilte Künstliche Intelligenz und kooperatives Arbeiten. Informatik-Fachberichte, vol 291. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-76980-1_43
Download citation
DOI: https://doi.org/10.1007/978-3-642-76980-1_43
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-54617-7
Online ISBN: 978-3-642-76980-1
eBook Packages: Springer Book Archive