Supporting Figures and Tables for Oil Contamination Prediction from Sparse Microbial Data

Supporting information for a study on predicting oil contamination using microbial community compositions. The dataset, authored by Tong Gao and last updated on March 19, 2026, is a 1.6 MB ZIP file containing figures and tables. It likely relates to a model that used a variational autoencoder and 10 bacterial features to mitigate data scarcity.

Use Cases

• Probe model robustness based on the described stress testing and principal component analysis
• Improve predictive performance for oil contamination based on the described use of 10 bacterial features
• Analyze microbial community compositions based on the described sparse microbial data
• Mitigate data scarcity in environmental modeling based on the described use of a variational autoencoder

Strengths

• Released under a permissive CC-BY-4.0 license
• Last updated on March 19, 2026, indicating recent maintenance
• Small size (1.6 MB) facilitates quick download and inspection

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download
• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment

Provenance

Source

figshare, authored by Tong Gao

Freshness

Last updated 2026-03-19 17:38:31