Closed-loop fully-automated frameworks for accelerating materials discovery
How much can a combination of sequential learning and closed-loop automation accelerate materials discovery?
Quantifying the Performance of Machine Learning Models in Materials Discovery
Our findings illustrate the importance of designing a modeling and discovery workflow which is highly tailored toward the specific problem of interest. In ML and SL for materials discovery, one size does not fit all, traditional error metrics may not be a good predictor of success, and the specifics of the design challenge have important implications for the optimal configuration of materials discovery strategies.
A new open database of corrosion-resistant alloys
Using machine learning to create a new open database of corrosion-resistant alloys.
Addressing challenges in high-temperature alloy design
High-temperature alloy design is challenging, but crucial to aeronautical applications. Citrine is working on a project to improve the representation of alloys data and develop techniques to to accurately predict properties outside of the range of input data.
Aggregating data to accelerate alloy design
Extracting, cleaning and analyzing and visualizing MPEA data.
Accelerating electrochemical material discovery for sustainable chemical manufacturing
New electrochemical reactions are needed to reduce the carbon footprint of chemical production. Use AI to find new catalysts.
Does machine learning work in materials?
Summary of peer reviewed, papers showing experimentally verified predictions of material properties.
Quantifying Uncertainties in the Materials Digital Twin
Citrine has quantified uncertainty in DFT property calculations as part of a project funded by the US Department of Energy and with collaborators from Olin College and MolSSI.