Learning Analytical Knowledge About VLSI-Design from Observation
Abstract
The LIMES system presented here learns estimation knowledge for the early stages of VLSI design - a novel and promising application. It integrates different machine learning methods based on a concept of similarity of module specifications and realizations already existing in the domain. LIMES evaluates and hierarchically classifies examples of IC modules by observing a module generator. This procedure generates a graph of classes used for estimation. These classes are generalized or specialized by means of inductive learning methods driven by unsuccessful estimations. Active experimentation supplies additional dedicated examples that support reasoning about the improvement and enlargement of the knowledge base.
Cite
Text
Herrmann. "Learning Analytical Knowledge About VLSI-Design from Observation." International Conference on Machine Learning, 1991. doi:10.1016/B978-1-55860-200-7.50124-0Markdown
[Herrmann. "Learning Analytical Knowledge About VLSI-Design from Observation." International Conference on Machine Learning, 1991.](https://mlanthology.org/icml/1991/herrmann1991icml-learning/) doi:10.1016/B978-1-55860-200-7.50124-0BibTeX
@inproceedings{herrmann1991icml-learning,
title = {{Learning Analytical Knowledge About VLSI-Design from Observation}},
author = {Herrmann, Jürgen},
booktitle = {International Conference on Machine Learning},
year = {1991},
pages = {610-614},
doi = {10.1016/B978-1-55860-200-7.50124-0},
url = {https://mlanthology.org/icml/1991/herrmann1991icml-learning/}
}