Machine-learning-assisted materials discovery using failed experiments

By Paul Raccuglia, Katherine C. Elbert, Philip Adler, Casey Falk, Malia B. Wenny, Aurelio Mollo, M. Zeller, Sorelle A. Friedler, Joshua Schrier, A. Norquist. 2016

1,511 citations recorded by the source corpus.

Subfields: Inorganic Chemistry, Materials Chemistry

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Inorganic Chemistry

1. Logan T. Ward
2. Alán Aspuru-Guzik
3. Zhen Zhou
4. Chris Wolverton
5. Pascal Friederich
6. Florian Hase

In Materials Chemistry

1. José Miguel Hernández-Lobato
2. Alán Aspuru-Guzik
3. Gisbert Schneider
4. Rafael Gómez‐Bombarelli
5. Tommi S. Jaakkola
6. Francesca Grisoni
7. Kenneth Atz
8. Regina Barzilay

Citations in the corpus

• Auditing black-box models for indirect influence (2016, background): “The top ranked decision tree features additionally include the averagemolecular polarizability for the organic components, whichwas previously hypothesized as important to synthesis of templated vanadium selenites explored via this data set [24].”
• Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules (2016, background): “Figures 13-15 ) Bayesian optimization of logP Having shown the ability of a VAE to generate a continuous, reversible encoding of molecular structures, we proceeded to test its potential to discover new molecules with desired properties.”
• How do Humans Understand Explanations from Machine Learning Systems? An Evaluation of the Human-Interpretability of Explanation (2018, background): “…has lead to an increased attention to interpretable machine learning systems, with applications including understanding notifications on mobile devices [Mehrotra et al., 2017, Wang et al., 2016], calculating stroke risk [Letham et al., 2015], and designing materials [Raccuglia et al., 2016].”
• All-optical neural network with nonlinear activation functions (2019, background): “ANNs also show great potential in scientific researches [3–6], especially in discovering new materials [7], classifying phases of matter [8, 9], representing variational wave functions [10], accelerating Monte Carlo simulations [11, 12], etc.”
• Integrating Physics-Based Modeling with Machine Learning: A Survey (2020, background): “There is also increasing interest in the inverse design of materials using ML, where desired target properties of materials are used as input to the model to identify atomic structures that exhibit such properties [137, 151, 202, 226].”