Unsupervised Regression with Applications to Nonlinear System Identification
- Creators
- Rahimi, Ali
- Recht, Ben
Abstract
We derive a cost functional for estimating the relationship between high-dimensional observations and the low-dimensional process that generated them with no input-output examples. Limiting our search to invertible observation functions confers numerous benefits, including a compact representation and no suboptimal local minima. Our approximation algorithms for optimizing this cost functional are fast and give diagnostic bounds on the quality of their solution. Our method can be viewed as a manifold learning algorithm that utilizes a prior on the low-dimensional manifold coordinates. The benefits of taking advantage of such priors in manifold learning and searching for the inverse observation functions in system identification are demonstrated empirically by learning to track moving targets from raw measurements in a sensor network setting and in an RFID tracking experiment.
Additional Information
© 2007 Massachusetts Institute of Technology.Attached Files
Published - 3039-unsupervised-regression-with-applications-to-nonlinear-system-identification.pdf
Files
Name | Size | Download all |
---|---|---|
md5:662132655a5e374730457e1886e83c70
|
1.9 MB | Preview Download |
Additional details
- Eprint ID
- 65342
- Resolver ID
- CaltechAUTHORS:20160314-160139280
- Created
-
2016-03-30Created from EPrint's datestamp field
- Updated
-
2019-10-03Created from EPrint's last_modified field
- Series Name
- Advances in Neural Information Processing Systems
- Series Volume or Issue Number
- 19