Search research articles

Related Concept Videos

Predicting Molecular Geometry

Predicting Molecular Geometry

VSEPR Theory for Determination of Electron Pair Geometries

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Multiscale Modeling of the RESET Sweep in a Single-Layer MoS<sub>2</sub> Atomristor Using Density Functional Theory.

ACS omega·2026

Same author

Finite-Temperature Evolution of Frenkel Defects in Hybrid Perovskites: Healing and Lead-Methylammonium Antisite Pairs.

ACS applied materials & interfaces·2026

Same author

Computational Workflow to Unravel the Structural Dynamics of Supramolecular Metallacages in Solution.

Journal of chemical theory and computation·2025

Same author

Synthesizer: Chemistry-Aware Machine Learning for Precision Control of Nanocrystal Growth.

Advanced materials (Deerfield Beach, Fla.)·2025

Same author

Trace Adsorptive Removal of PFAS from Water by Optimizing the UiO-66 MOF Interface.

Advanced materials (Deerfield Beach, Fla.)·2024

Same author

Conductive filament distribution in nano-scale electrochemical metallization cells.

Nanoscale·2024

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Nov 4, 2025

Monovalent Cation Doping of CH3NH3PbI3 for Efficient Perovskite Solar Cells

Monovalent Cation Doping of CH3NH3PbI3 for Efficient Perovskite Solar Cells

Published on: March 19, 2017

Global Property Prediction: A Benchmark Study on Open-Source, Perovskite-like Datasets.

Felix Mayr¹, Alessio Gagliardi¹

¹Department of Electrical and Computer Engineering, Technical University of Munich, Munich 80333, Germany.

|May 31, 2021

Summary

Machine learning models for perovskite materials struggle to generalize across different databases. Common metrics and fingerprints are database-dependent, highlighting the need for improved methods in materials discovery.

More Related Videos

Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance

Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance

Published on: February 27, 2017

Flash Infrared Annealing for Perovskite Solar Cell Processing

Flash Infrared Annealing for Perovskite Solar Cell Processing

Published on: February 3, 2021

Related Experiment Videos

Last Updated: Nov 4, 2025

Monovalent Cation Doping of CH3NH3PbI3 for Efficient Perovskite Solar Cells

Monovalent Cation Doping of CH3NH3PbI3 for Efficient Perovskite Solar Cells

Published on: March 19, 2017

Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance

Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance

Published on: February 27, 2017

Flash Infrared Annealing for Perovskite Solar Cell Processing

Flash Infrared Annealing for Perovskite Solar Cell Processing

Published on: February 3, 2021

Area of Science:

Materials Science
Computational Chemistry
Machine Learning

Background:

High-throughput screening generates vast data for novel materials like perovskites.
Predicting properties such as band gaps and energies is crucial for photovoltaic applications.

Purpose of the Study:

To comprehensively compare structural fingerprint-based machine learning models.
To evaluate model performance across seven open-source perovskite-like material databases.
To assess the database dependency of common metrics and fingerprints.

Main Methods:

Utilized seven open-source perovskite-like material databases.
Applied various structural fingerprint-based machine learning models, including graph neural networks.
Investigated variance selection and autoencoders for fingerprint dimensionality reduction.

Main Results:

No single machine learning method demonstrated consistent performance across all tested databases.
Commonly used performance metrics were found to be highly database-dependent.
Dimensionality reduction techniques revealed that models often utilize only a subset of the available fingerprint space.

Conclusions:

Current machine learning approaches for perovskite materials exhibit limited generalizability across diverse datasets.
The choice of database and evaluation metrics significantly impacts perceived model performance.
Further research is needed to develop more robust and transferable machine learning models for materials discovery.