Exploring effective charge in electromigration using machine learning

Yu Chen Liu, Benjamin Afflerbach, Ryan Jacobs, Shih Kang Lin, Dane Morgan

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The effective charge of an element is a parameter characterizing the electromigration effect, which can determine the reliability of interconnection in electronic technologies. In this work, machine learning approaches were employed to model the effective charge (z) as a linear function of physically meaningful elemental properties. Average fivefold (leave-out-alloy-group) cross-validation yielded root-mean-square-error divided by whole data set standard deviation (RMSE/σ) values of 0.37 ± 0.01 (0.22 ± 0.18), respectively, and R2 values of 0.86. Extrapolation to z∗ of totally new alloys showed limited but potentially useful predictive ability. The model was used in predicting z∗ for technologically relevant host-impurity pairs.

Original languageEnglish
Pages (from-to)567-575
Number of pages9
JournalMRS Communications
Volume9
Issue number2
DOIs
Publication statusPublished - 2019 Jun 1

Fingerprint

Electromigration
Learning systems
Extrapolation
Mean square error
Impurities

All Science Journal Classification (ASJC) codes

  • Materials Science(all)

Cite this

Liu, Yu Chen ; Afflerbach, Benjamin ; Jacobs, Ryan ; Lin, Shih Kang ; Morgan, Dane. / Exploring effective charge in electromigration using machine learning. In: MRS Communications. 2019 ; Vol. 9, No. 2. pp. 567-575.
@article{0f92a52011ae4c139e3960d4c9f88234,
title = "Exploring effective charge in electromigration using machine learning",
abstract = "The effective charge of an element is a parameter characterizing the electromigration effect, which can determine the reliability of interconnection in electronic technologies. In this work, machine learning approaches were employed to model the effective charge (z∗) as a linear function of physically meaningful elemental properties. Average fivefold (leave-out-alloy-group) cross-validation yielded root-mean-square-error divided by whole data set standard deviation (RMSE/σ) values of 0.37 ± 0.01 (0.22 ± 0.18), respectively, and R2 values of 0.86. Extrapolation to z∗ of totally new alloys showed limited but potentially useful predictive ability. The model was used in predicting z∗ for technologically relevant host-impurity pairs.",
author = "Liu, {Yu Chen} and Benjamin Afflerbach and Ryan Jacobs and Lin, {Shih Kang} and Dane Morgan",
year = "2019",
month = "6",
day = "1",
doi = "10.1557/mrc.2019.63",
language = "English",
volume = "9",
pages = "567--575",
journal = "MRS Communications",
issn = "2159-6859",
publisher = "Cambridge University Press",
number = "2",

}

Exploring effective charge in electromigration using machine learning. / Liu, Yu Chen; Afflerbach, Benjamin; Jacobs, Ryan; Lin, Shih Kang; Morgan, Dane.

In: MRS Communications, Vol. 9, No. 2, 01.06.2019, p. 567-575.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Exploring effective charge in electromigration using machine learning

AU - Liu, Yu Chen

AU - Afflerbach, Benjamin

AU - Jacobs, Ryan

AU - Lin, Shih Kang

AU - Morgan, Dane

PY - 2019/6/1

Y1 - 2019/6/1

N2 - The effective charge of an element is a parameter characterizing the electromigration effect, which can determine the reliability of interconnection in electronic technologies. In this work, machine learning approaches were employed to model the effective charge (z∗) as a linear function of physically meaningful elemental properties. Average fivefold (leave-out-alloy-group) cross-validation yielded root-mean-square-error divided by whole data set standard deviation (RMSE/σ) values of 0.37 ± 0.01 (0.22 ± 0.18), respectively, and R2 values of 0.86. Extrapolation to z∗ of totally new alloys showed limited but potentially useful predictive ability. The model was used in predicting z∗ for technologically relevant host-impurity pairs.

AB - The effective charge of an element is a parameter characterizing the electromigration effect, which can determine the reliability of interconnection in electronic technologies. In this work, machine learning approaches were employed to model the effective charge (z∗) as a linear function of physically meaningful elemental properties. Average fivefold (leave-out-alloy-group) cross-validation yielded root-mean-square-error divided by whole data set standard deviation (RMSE/σ) values of 0.37 ± 0.01 (0.22 ± 0.18), respectively, and R2 values of 0.86. Extrapolation to z∗ of totally new alloys showed limited but potentially useful predictive ability. The model was used in predicting z∗ for technologically relevant host-impurity pairs.

UR - http://www.scopus.com/inward/record.url?scp=85066884749&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066884749&partnerID=8YFLogxK

U2 - 10.1557/mrc.2019.63

DO - 10.1557/mrc.2019.63

M3 - Article

AN - SCOPUS:85066884749

VL - 9

SP - 567

EP - 575

JO - MRS Communications

JF - MRS Communications

SN - 2159-6859

IS - 2

ER -