DOI | Resolve DOI: https://doi.org/10.1109/PN56061.2022.9908251 |
---|
Author | Search for: Grinberg, Yuri1; Search for: Al-Digeil, Muhammad1; Search for: Kamandar Dezfouli, Mohsen2; Search for: Melati, Daniele2; Search for: Schmid, Jens H.2; Search for: Cheben, Pavel2; Search for: Janz, Siegfried2; Search for: Xu, Danxia2 |
---|
Affiliation | - National Research Council of Canada. Digital Technologies
- National Research Council of Canada. Advanced Electronics and Photonics
|
---|
Funder | Search for: National Research Council of Canada |
---|
Format | Text, Article |
---|
Conference | 2022 Photonics North (PN), May 24-26, 2022, Niagara Falls, Ontario, Canada |
---|
Subject | dimensionality reduction; autoencoder; small data; nanophotonic design; silicon photonics; machine learning; dimensionality reduction; neural networks; photonics; principal component analysis |
---|
Abstract | Efficient exploration of high-dimensional parameter space is essential in modern photonic component design. Linear dimensionality reduction such as principal component analysis has proven useful in identifying lower dimensional subspace of interest in several design problems. Yet such subspaces often exhibit curvature reflecting nonlinear relationships between design parameters. For such systems linear dimensionality reduction methods can be suboptimal. We discuss how an appropriate architecture for an autoencoder neural network along with a numerically robust initialization, show improved performance compared to linear methods even in low data regimes, which are typical for photonic design problems. |
---|
Publication date | 2022-05-24 |
---|
Publisher | IEEE |
---|
In | |
---|
Language | English |
---|
Peer reviewed | Yes |
---|
Export citation | Export as RIS |
---|
Report a correction | Report a correction (opens in a new tab) |
---|
Record identifier | c548520b-92e2-48da-af32-fc26a9d79d1c |
---|
Record created | 2022-10-25 |
---|
Record modified | 2023-03-16 |
---|