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

In order to make accurate predictions of material properties, currentmachine-learning approaches generally require large amounts of data, which areoften not available in practice. In this work, an all-round framework ispresented which relies on a feedforward neural network, the selection ofphysically-meaningful features and, when applicable, joint-learning. Next tobeing faster in terms of training time, this approach is shown to outperformcurrent graph-network models on small datasets. In particular, the vibrationalentropy at 305 K of crystals is predicted with a mean absolute test error of0.009 meV/K/atom (four times lower than previous studies). Furthermore,joint-learning reduces the test error compared to single-target learning andenables the prediction of multiple properties at once, such as temperaturefunctions. Finally, the selection algorithm highlights the most importantfeatures and thus helps understanding the underlying physics.

MODNet -- accurate and interpretable property predictions for limited materials datasets by feature selection and joint-learning