Alternative robust statistical methods to reduce parameters uncertainty: application to scatterometry

J. Hazart; F. Sarrazy; R. Buyssou; C. Dezauzier

doi:10.1117/12.889493

23 May 2011 Alternative robust statistical methods to reduce parameters uncertainty: application to scatterometry

J. Hazart, F. Sarrazy, R. Buyssou, C. Dezauzier

Proceedings Volume 8083, Modeling Aspects in Optical Metrology III; 80830J (2011) https://doi.org/10.1117/12.889493
Event: SPIE Optical Metrology, 2011, Munich, Germany

Abstract

The determination of interesting parameters is very often performed through curves fitting, like in scatterometry for example. Classically, the problem is solved by means of least-squares. Although optimal in an ideal context of Normal errors, this assumption is not valid mainly because of modeling and calibration errors, and induce some uncontrolled biases. In this paper, we propose radically different principles of fitting techniques, based on an Entropy criterion instead of the Maximum Likelihood Principle. Combining simple implementation and high performance, we show that this technique is optimal for pure Normal noise and dramatically reduces bias on parameters for corrupted data, outperforming conventional robust M-Estimators. We also provide tests on real scatterometry samples which demonstrate a bias reduction of a factor 4.

Citation Download Citation

J. Hazart, F. Sarrazy, R. Buyssou, and C. Dezauzier "Alternative robust statistical methods to reduce parameters uncertainty: application to scatterometry", Proc. SPIE 8083, Modeling Aspects in Optical Metrology III, 80830J (23 May 2011); https://doi.org/10.1117/12.889493

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available