English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Material scientists are increasingly adopting the use of machine learning(ML) for making potentially important decisions, such as, discovery,development, optimization, synthesis and characterization of materials.However, despite ML's impressive performance in commercial applications,several unique challenges exist when applying ML in materials scienceapplications. In such a context, the contributions of this work are twofold.First, we identify common pitfalls of existing ML techniques when learning fromunderrepresented/imbalanced material data. Specifically, we show that withimbalanced data, standard methods for assessing quality of ML models break downand lead to misleading conclusions. Furthermore, we found that the model's ownconfidence score cannot be trusted and model introspection methods (usingsimpler models) do not help as they result in loss of predictive performance(reliability-explainability trade-off). Second, to overcome these challenges,we propose a general-purpose explainable and reliable machine-learningframework. Specifically, we propose a novel pipeline that employs an ensembleof simpler models to reliably predict material properties. We also propose atransfer learning technique and show that the performance loss due to models'simplicity can be overcome by exploiting correlations among different materialproperties. A new evaluation metric and a trust score to better quantify theconfidence in the predictions are also proposed. To improve theinterpretability, we add a rationale generator component to our framework whichprovides both model-level and decision-level explanations. Finally, wedemonstrate the versatility of our technique on two applications: 1) predictingproperties of crystalline compounds, and 2) identifying novel potentiallystable solar cell materials.

Reliable and Explainable Machine Learning Methods for Accelerated Material Discovery