scholarly journals Opportunities and Challenges for Machine Learning in Materials Science

2020 ◽  
Vol 50 (1) ◽  
pp. 71-103
Author(s):  
Dane Morgan ◽  
Ryan Jacobs
Keyword(s):  
Machine Learning ◽  
Best Practices ◽  
Materials Science ◽  
Molecular Simulations ◽  
Learning Models ◽  
Novel Materials ◽  
Domain Of Applicability ◽  
Rapid Changes ◽  
To Come ◽  
Machine Learning Models

Advances in machine learning have impacted myriad areas of materials science, such as the discovery of novel materials and the improvement of molecular simulations, with likely many more important developments to come. Given the rapid changes in this field, it is challenging to understand both the breadth of opportunities and the best practices for their use. In this review, we address aspects of both problems by providing an overview of the areas in which machine learning has recently had significant impact in materials science, and then we provide a more detailed discussion on determining the accuracy and domain of applicability of some common types of machine learning models. Finally, we discuss some opportunities and challenges for the materials community to fully utilize the capabilities of machine learning.

Machine Learning Models for Lipophilicity and Their Domain of Applicability

2007 ◽  
Vol 4 (4) ◽  
pp. 524-538 ◽  
Author(s):  
Timon Schroeter ◽  
Anton Schwaighofer ◽  
Sebastian Mika ◽  
Antonius Ter Laak ◽  
Detlev Suelzle ◽  
...  
Keyword(s):  
Machine Learning ◽  
Learning Models ◽  
Domain Of Applicability ◽  
Machine Learning Models

scholarly journals Identifying Domains of Applicability of Machine Learning Models for Materials Science

2019 ◽  
Author(s):  
Christopher Sutton ◽  
Mario Boley ◽  
Luca M. Ghiringhelli ◽  
Matthias Rupp ◽  
Jilles Vreeken ◽  
...  
Keyword(s):  
Machine Learning ◽  
Ground State ◽  
Materials Science ◽  
Many Body ◽  
Simple Set ◽  
Domain Of Applicability ◽  
N Gram ◽  
The Many ◽  
Error Statistic ◽  
Machine Learning Models

We present an extension to the usual machine learning process that allows for the identification of the domain of applicability of a fitted model, i.e., the region in its domain where it performs most accurately. This approach is applied to several vastly different but commonly used materials representations (namely the n-gram approach, SOAP, and the many body tenor representation), which are practically indistinguishable based on performance using a single error statistic. Moreover, these models appear unsatisfactory for screening applications as they fail to reliably identify the ground state polymorphs. When applying our newly developed analysis for each of the models, we can identify the domain of applicability for each model according to a simple set of interpretable conditions. We show that identification of the domain of applicability in the prediction of the formation energy enables a more accurate ground-state search - a crucial step for the discovery of novel materials.

scholarly journals Identifying domains of applicability of machine learning models for materials science

2020 ◽  
Vol 11 (1) ◽  
Author(s):  
Christopher Sutton ◽  
Mario Boley ◽  
Luca M. Ghiringhelli ◽  
Matthias Rupp ◽  
Jilles Vreeken ◽  
...  
Keyword(s):  
Machine Learning ◽  
Materials Science ◽  
Learning Models ◽  
Machine Learning Models

scholarly journals Identifying Domains of Applicability of Machine Learning Models for Materials Science

2019 ◽  
Author(s):  
Christopher Sutton ◽  
Mario Boley ◽  
Luca M. Ghiringhelli ◽  
Matthias Rupp ◽  
Jilles Vreek ◽  
...  
Keyword(s):  
Machine Learning ◽  
Ground State ◽  
Materials Science ◽  
Many Body ◽  
Simple Set ◽  
Domain Of Applicability ◽  
N Gram ◽  
The Many ◽  
Error Statistic ◽  
Machine Learning Models

We present an extension to the usual machine learning process that allows for the identification of the domain of applicability of a fitted model, i.e., the region in its domain where it performs most accurately. This approach is applied to several vastly different but commonly used materials representations (namely the n-gram approach, SOAP, and the many body tenor representation), which are practically indistinguishable based on performance using a single error statistic. Moreover, these models appear unsatisfactory for screening applications as they fail to reliably identify the ground state polymorphs. When applying our newly developed analysis for each of the models, we can identify the domain of applicability for each model according to a simple set of interpretable conditions. We show that identification of the domain of applicability in the prediction of the formation energy enables a more accurate ground-state search - a crucial step for the discovery of novel materials.

scholarly journals Identifying Domains of Applicability of Machine Learning Models for Materials Science

2019 ◽  
Author(s):  
Christopher Sutton ◽  
Mario Boley ◽  
Luca M. Ghiringhelli ◽  
Matthias Rupp ◽  
Jilles Vreeken ◽  
...  
Keyword(s):  
Machine Learning ◽  
Ground State ◽  
Materials Science ◽  
Many Body ◽  
Simple Set ◽  
Domain Of Applicability ◽  
N Gram ◽  
The Many ◽  
Error Statistic ◽  
Machine Learning Models

We present an extension to the usual machine learning process that allows for the identification of the domain of applicability of a fitted model, i.e., the region in its domain where it performs most accurately. This approach is applied to several vastly different but commonly used materials representations (namely the n-gram approach, SOAP, and the many body tenor representation), which are practically indistinguishable based on performance using a single error statistic. Moreover, these models appear unsatisfactory for screening applications as they fail to reliably identify the ground state polymorphs. When applying our newly developed analysis for each of the models, we can identify the domain of applicability for each model according to a simple set of interpretable conditions. We show that identification of the domain of applicability in the prediction of the formation energy enables a more accurate ground-state search - a crucial step for the discovery of novel materials.

scholarly journals Going beyond simplicity: Using machine learning to predict belief in conspiracy theories

2021 ◽  
Author(s):  
Nils Brandenstein
Keyword(s):  
Machine Learning ◽  
Psychological Factors ◽  
Theoretical Explanation ◽  
Theory Building ◽  
Learning Models ◽  
Current Analysis ◽  
Conspiracy Theories ◽  
The Past ◽  
To Come ◽  
Machine Learning Models

Public and scientific interest in why people believe in conspiracy theories (CT) surged in the past years. To come up with a theoretical explanation, researchers investigated relationships of CT belief with psychological factors such as political attitudes, emotions or personality (van Prooijen & Douglas, 2018). However, recent studies put the robustness of these relationships into question (e.g., Stojanov & Halberstadt, 2020). In this study, the analysis of a representative dataset with 2025 adults uncovered that the simplicity of the current analysis routine, exhibiting high sample-specificity and neglecting complex associations of psychological factors and belief in CTs, may obscure these relationships. Further, poor replicability of CT belief associations can be detected and remedied by using a prediction-based modeling approach and machine learning models, which proposes a timely shift in the field’s analysis routine. Conceptual and theoretical implications for CT belief research and theory building are derived.

Improving XGBoost with Imagination Sampling

2020 ◽  
Vol 2 (1) ◽  
pp. 3-6
Author(s):  
Eric Holloway
Keyword(s):  
Machine Learning ◽  
General System ◽  
Learning Models ◽  
Starting Point ◽  
Machine Learning Models

Imagination Sampling is the usage of a person as an oracle for generating or improving machine learning models. Previous work demonstrated a general system for using Imagination Sampling for obtaining multibox models. Here, the possibility of importing such models as the starting point for further automatic enhancement is explored.

Sign in / Sign up

Export Citation Format

Share Document