Better, Faster, and Less Biased Machine Learning: Electromechanical Switching in PZT Thin Films

This repository contains the data and code used in the corresponding study.

Note: This study is based on previously reported data[1,2], which is contained in Data.mat and loop_1.mat. Data.mat and loop_1.mat are uploaded here for ease of access and are completely unaltered from how they were provided by the original authors. The data files are available from the original authors on Zenodo[3,4].

The abstract for this work is copied here:

"Machine learning techniques are more and more often applied to the analysis of complex behaviors in materials research. Frequently used to identify fundamental behaviors within large and multidimensional datasets, these techniques are strictly based on mathematical models. Thus, without inherent physical or chemical meaning or constraints, they are prone to biased interpretation.

Here, we demonstrate that, through physical insights and careful data handling, the interpretability of machine learning results in materials science, and specifically materials functionalities, can be vastly improved. The use of techniques such as dimensional stacking can provide the much needed physical and chemical constraints, while proper understanding of the assumptions imposed by model parameters can help avoid over-interpretation. These concepts are applied to recently reported ferroelectric switching experiments in PZT thin films. Through systematic analysis and introduction of physical constraints, we argue that the behaviors present are not necessarily due to exotic mechanisms, as previously suggested, but rather well described by classical ferroelectric switching superimposed by non-ferroelectric phenomena, such as electrochemical deformation, electrostatic interactions or charge injection."

[1] Agar et. al. J. C. Agar, Y. Cao, B. Naul, S. Pandya, S. van der Walt, A. I. Luo, J. T. Maher, N. Balke,S. Jesse, S. V. Kalinin, R. K. Vasudevan,  and L. W. Martin, Advanced Materials30, 1800701(2018).
[2] J. C. Agar, B. Naul, S. Pandya, S. van der Walt, J. Maher, Y. Ren, L.-Q. Chen, S. V. Kalinin,R. K. Vasudevan, Y. Cao, J. S. Bloom,  and L. W. Martin, Nature Communications10, 4809(2019).
[3] J.  C.  Agar,  JAgar,  Y.  Cao,  B.  Naul,  S.  Pandya,  S.  van  der  Walt,  A.  I.  Luo,  J.  T.  Maher,N. Balke, S. Jesse, S. V. Kalinin, R. K. Vasudevan,  and L. W. C. Martin,  (2018), 10.5281/ZEN-ODO.1242656
[4] J.  C.  Agar,  B.  Naul,  S.  Pandya,  S.  van  der  Walt,  R.  Yao,  J.  Maher,  J.  Neaton,  S.  Kalinin,R. Vasudevan, Y. Cao, J. Bloom,  and L. Martin,  (2018), 10.5281/ZENODO.1478089