Training Data for Active Delta-Learning Models in Surface Hopping

Training data for machine learning models discussed in the manuscript 'Active Delta-Learning for Surface Hopping: Efficient Fitting of Potential Energy Surfaces'. The 1.0 GB dataset was authored by Mikołaj Martyka and last updated on April 20, 2026. Associated models are available in a linked GitHub repository.

Use Cases

• Training machine learning models for potential energy surfaces based on the active delta-learning method described.
• Benchmarking new algorithms for efficient fitting of quantum chemistry data.
• Reproducing or extending the results of the associated surface hopping manuscript.

Strengths

• Data volume is 1.0 GB, indicating a medium-scale training set.
• Data is directly linked to a peer-reviewed manuscript and associated open-source code repository.
• Released under a permissive CC-BY-4.0 license for reuse.

Limitations

• Row count and specific column definitions are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Mikołaj Martyka

Collection Method

Likely generated from quantum chemistry calculations for the referenced research.

Freshness

Last updated 2026-04-20 09:43:27; freshness should be verified.