More Like this

1. Applications and Techniques for Fast Machine Learning in Science

   https://doi.org/10.3389/fdata.2022.787421  

   Deiana, Allison McCarn; Tran, Nhan; Agar, Joshua; Blott, Michaela; Di Guglielmo, Giuseppe; Duarte, Javier; Harris, Philip; Hauck, Scott; Liu, Mia; Neubauer, Mark S.; et al (April 2022, Frontiers in Big Data)

   In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs. 

   more » « less
2. Is Community Relevance Enough? Civic and Science Identity Impact of Microbiology CUREs Focused on Community Environmental Justice

   https://doi.org/10.3389/fmicb.2020.578520  

   Adkins-Jablonsky, Sarah J.; Akscyn, Rob; Bennett, Brad C.; Roberts, Qutia; Morris, J. Jeffrey (December 2020, Frontiers in Microbiology)

   null (Ed.)

   Course-based undergraduate research experiences (CUREs) often involve a component where the outcomes of student research are broadly relevant to outside stakeholders. We wanted to see if building courses around an environmental justice issue relevant to the local community would impact students’ sense of civic engagement and appreciation of the relevance of scientific research to the community. In this quasi-experimental study, we assessed civic engagement and scientific identity gains ( N = 98) using pre- and post-semester surveys and open-ended interview responses in three different CUREs taught simultaneously at three different universities. All three CURES were focused on an environmental heavy metal pollution issue predominantly affecting African–Americans in Birmingham, Alabama. While we found increases in students’ sense of science efficacy and identity, our team was unable to detect meaningful changes in civic engagement levels, all of which were initially quite high. However, interviews suggested that students were motivated to do well in their research because the project was of interest to outside stakeholders. Our observations suggest that rather than directly influencing students’ civic engagement, the “broadly relevant” component of our CUREs engaged their pre-existing high levels of engagement to increase their engagement with the material, possibly influencing gains in science efficacy and science identity. Our observations are consistent with broader community relevance being an important component of CURE success, but do not support our initial hypothesis that CURE participation would influence students’ attitudes toward the civic importance of science. 

   more » « less
3. Democratizing machine learning in chemistry with community-engaged test sets

   https://doi.org/10.1039/D5DD00424A  

   Wu, Jason L; Friday, David M; Hwang, Changhyun; Yi, Seungjoo; Torres-Flores, Tiara C; Burke, Martin D; Diao, Ying; Schroeder, Charles M; Jackson, Nicholas E (January 2026, Digital Discovery)

   Machine learning (ML) is increasingly central to chemical discovery, yet most efforts remain confined to distributed and isolated research groups, limiting external validation and community engagement. 

   more » « less
4. An efficient way of harvesting mechanoluminescent light: Flexible pressure sensor application

   https://doi.org/10.1016/j.jlumin.2023.119895  

   Shohag, Md Abu; Adams, Geoffrey; Eze, Vincent; Carani, Lucas Braga; Ichite, Tony; Okoli, Okenwa (August 2023, Journal of Luminescence)

   Mechanoluminescent (ML) materials are used for fabricating sensors and other devices such as artificial skin, colorful displays, and energy harvesting devices. However, a key challenge in developing ML-based sensors is the ability to effectively capture and efficiently transmit ML light from the sensing location. Here we report a flexible and sensitive thin film pressure sensor, created using a novel combination of ML material and perovskite. In this work, we adopted a simple lateral type design of a thin pressure sensor primarily consisting of (i) a sensing layer of copper-doped zinc sulfide (ZnS:Cu)/polydimethylsiloxane (PDMS) composite and (ii) a light absorbing layer of perovskite. The mixed halide perovskite, a light absorbing material, fully absorbs the green light emitted from ZnS:Cu. The sensor demonstrated consistent signal output under the mechanical bending test. A thin encapsulation layer of PMMA on the perovskite layer prevents moisture inclusion. This innovative technique of utilizing integrated thin perovskite to efficiently harvest ML light has the potential to open up new avenues for advanced research in ML-perovskite-based sensor systems. 

   more » « less
5. Learning Molecular Mixture Property Using Chemistry-Aware Graph Neural Network

   https://doi.org/10.1103/PRXEnergy.3.023006  

   Zhang, Hengrui; Lai, Tianxing; Chen, Jie; Manthiram, Arumugam; Rondinelli, James M; Chen, Wei (June 2024, PRX Energy)

   Recent advances in machine learning (ML) are expediting materials discovery and design. One significant challenge facing ML for materials is the expansive combinatorial space of potential materials formed by diverse constituents and their flexible configurations. This complexity is particularly evident in molecular mixtures, a frequently explored space for materials, such as battery electrolytes. Owing to the complex structures of molecules and the sequence-independent nature of mixtures, conventional ML methods have difficulties in modeling such systems. Here, we present MolSets, a specialized ML model for molecular mixtures, to overcome the difficulties. Representing individual molecules as graphs and their mixture as a set, MolSets leverages a graph neural network and the deep sets architecture to extract information at the molecular level and aggregate it at the mixture level, thus addressing local complexity while retaining global flexibility. We demonstrate the efficacy of MolSets in predicting the conductivity of lithium battery electrolytes and highlight its benefits in the virtual screening of the combinatorial chemical space. Published by the American Physical Society2024 

   more » « less