Neural Network Exploration Using Optimal Experiment Design

Abstract

Consider the problem of learning input/output mappings through exploration, e.g. learning the kinematics or dynamics of a robotic manipulator. If actions are expensive and computation is cheap, then we should explore by selecting a trajectory through the in(cid:173) put space which gives us the most amount of information in the fewest number of steps. I discuss how results from the field of opti(cid:173) mal experiment design may be used to guide such exploration, and demonstrate its use on a simple kinematics problem.

Cite

Text

Cohn. "Neural Network Exploration Using Optimal Experiment Design." Neural Information Processing Systems, 1993.

Markdown

[Cohn. "Neural Network Exploration Using Optimal Experiment Design." Neural Information Processing Systems, 1993.](https://mlanthology.org/neurips/1993/cohn1993neurips-neural/)

BibTeX

@inproceedings{cohn1993neurips-neural,
  title     = {{Neural Network Exploration Using Optimal Experiment Design}},
  author    = {Cohn, David A.},
  booktitle = {Neural Information Processing Systems},
  year      = {1993},
  pages     = {679-686},
  url       = {https://mlanthology.org/neurips/1993/cohn1993neurips-neural/}
}