A5055766485- Machine Learning in Materials Science
- X-ray Diffraction in Crystallography
- Perovskite Materials and Applications
- High Temperature Alloys and Creep
- Conducting polymers and applications
- nanoparticles nucleation surface interactions
- Microstructure and Mechanical Properties of Steels
- Electrocatalysts for Energy Conversion
- Advanced Materials Characterization Techniques
- Magnetic properties of thin films
- Electronic and Structural Properties of Oxides
- Crystallization and Solubility Studies
- Fusion materials and technologies
- Organic Electronics and Photovoltaics
- Theoretical and Computational Physics
- Machine Learning and Algorithms
- Magnetic Properties and Applications
- Fuel Cells and Related Materials
- Computational Drug Discovery Methods
- Advanced Chemical Physics Studies
- Blood Pressure and Hypertension Studies
- Titanium Alloys Microstructure and Properties
- High Entropy Alloys Studies
- Machine Learning and Data Classification
- Advanced battery technologies research
University of Toronto
2022-2025
Xuzhou Medical College
2023-2024
University of New Brunswick
2023-2024
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
2019-2024
CEA Paris-Saclay
2019-2024
Université Paris-Saclay
2019-2024
Weatherford College
2024
Child Trends
2024
Sun Yat-sen University
2020-2024
National Institute of Metrology
2023
Abstract Recent advances in machine learning (ML) have led to substantial performance improvement material database benchmarks, but an excellent benchmark score may not imply good generalization performance. Here we show that ML models trained on Materials Project 2018 can severely degraded new compounds 2021 due the distribution shift. We discuss how foresee issue with a few simple tools. Firstly, uniform manifold approximation and projection (UMAP) be used investigate relation between...
Extensive efforts to gather materials data have largely overlooked potential redundancy. In this study, we present evidence of a significant degree redundancy across multiple large datasets for various material properties, by revealing that up 95% can be safely removed from machine learning training with little impact on in-distribution prediction performance. The redundant is related over-represented types and does not mitigate the severe performance degradation out-of-distribution samples....
Abstract Scientific machine learning (ML) aims to develop generalizable models, yet assessments of generalizability often rely on heuristics. Here, we demonstrate in the materials science setting that heuristic evaluations lead biased conclusions ML and benefits neural scaling, through out-of-distribution (OOD) tasks involving unseen chemistry or structural symmetries. Surprisingly, many good performance across including boosted trees. However, analysis representation space shows most test...
Abstract Lack of rigorous reproducibility and validation are significant hurdles for scientific development across many fields. Materials science, in particular, encompasses a variety experimental theoretical approaches that require careful benchmarking. Leaderboard efforts have been developed previously to mitigate these issues. However, comprehensive comparison benchmarking on an integrated platform with multiple data modalities perfect defect materials is still lacking. This work...
There is an increasing need to synthesize biocompatible nanofibers with excellent mechanical and electrical performance for electrochemical biomedical applications. Here we report a facile approach prepare electroactive flexible 3D nanostructured biomaterials high based on bacterial cellulose (BC) nanofibers. Our can coat BC poly(3,4-ethylenedioxythiophene) (PEDOT) by in situ interfacial polymerization controllable manner. The PEDOT coating thickness adjustable the monomer concentration or...
Moisture-assisted post-annealing was performed on carbon-electrode based planar perovskite solar cells so as to improve the hole-extraction process. It observed that, after being annealed at a relative humidity of 30% for 2 h, short-circuit current density, fill factor, and open circuit voltage were all improved, leading an improvement 21.75% power conversion efficiency [from 10.53 (±0.98)% 12.82 (±1.07)%, with optimized one 14.77% reverse scanning]. The transient photovoltage/photocurrent...
Electrochemical Impedance Spectroscopy (EIS) is a powerful tool for electrochemical analysis; however, its data can be challenging to interpret. Here, we introduce new open-source named AutoEIS that assists EIS analysis by automatically proposing statistically plausible equivalent circuit models (ECMs). does this without requiring an exhaustive mechanistic understanding of the systems. We demonstrate generalizability using it analyze datasets from three distinct systems, including thin-film...
Generalization performance of machine learning models: (upper panel) generalization from small ordered to large disordered structures (SQS); (lower low-order high-order systems.
Predicting atomic diffusion in concentrated magnetic systems is challenging due to thermal effects and complex magnetochemical interplay. We propose an efficient approach via kinetic Monte Carlo using ab initio parametrized models. demonstrate its accuracy the case of Fe-Ni alloys, where we successfully predict explain weak composition dependence coefficients a compensation distinct contributions their constituents. The diffusion-behavior difference between paramagnetic ground states...
Scientific machine learning (ML) endeavors to develop generalizable models with broad applicability. However, the assessment of generalizability is often based on heuristics. Here, we demonstrate in materials science setting that heuristics evaluations lead substantially biased conclusions ML and benefits neural scaling. We evaluate generalization performance over 700 out-of-distribution tasks features new chemistry or structural symmetry not present training data. Surprisingly, good found...
We investigate phase stability and vacancy formation in fcc Fe-Ni alloys over a broad composition-temperature range, via density functional theory parametrized effective interaction model, which includes explicitly spin chemical variables. On-lattice Monte Carlo simulations based on this model are used to predict the temperature evolution of magnetochemical phase. The experimental composition-dependent Curie order-disorder transition temperatures successfully predicted. point out significant...
Abstract Electrochemical Impedance Spectroscopy (EIS) is a crucial technique for assessing corrosion of metallic materials. The analysis EIS hinges on the selection an appropriate equivalent circuit model (ECM) that accurately characterizes system under study. In this work, we systematically examined applicability three commonly used ECMs across several typical material degradation scenarios. By applying Bayesian Inference to simulated data, assessed suitability these different conditions...
Abstract Large language models (LLMs) are increasingly being used in materials science. However, little attention has been given to benchmarking and standardized evaluation for LLM-based property prediction, which hinders progress. We present LLM4Mat-Bench, the largest benchmark date evaluating performance of LLMs predicting properties crystalline materials. LLM4Mat-Bench contains about 1.9M crystal structures total, collected from 10 publicly available data sources, 45 distinct properties....
The rising incidence of cardiovascular diseases (CVD) in the elderly highlights need for effective preventive strategies. Recent studies suggest that obesity, through metabolic factors, contributes to development CVD. This study aims explore how body roundness index (BRI) levels affect occurrence CVD using data from National Health and Nutrition Examination Survey (NHANES) (2003-2016), better understand role obesity prevention management. analyzed 3,584 NHANES participants over seven cycles...
We evaluate LLM performance and robustness for materials science Q&A property prediction. Prompt sensitivity mode collapse reveal reliability concerns, providing informed skepticism guiding more cautious adoption in scientific research.
Power conversion properties of perovskite solar cells are studied in the temperature range 310 K to 240 (and recovering back). As lowers down, fill factor (FF) decreases while open circuit voltage (VOC) increases case reverse scans (scanning from positive voltages negative ones). The decreased FF is ascribed increased resistance charge transport materials (both TiO2 and Spiro-OMeTAD) as well interfacial transfer resistance, VOC due retarded recombination which revealed by transient...