Experimental Data for SVM Modeling of CO2/CH4 Hydrate Inhibition with Ionic Liquids

946 experimental data points on CO2/CH4 binary mixture hydrate inhibition, used to train a Support Vector Machine regression model. The dataset was created by Dileep Kumar and last updated in June 2026 to systematically test seven input feature scaling techniques. It includes results for 63 ionic liquid inhibitors.

Use Cases

• Benchmarking feature scaling methods for regression models based on the seven tested preprocessing techniques.
• Training machine learning models to predict hydrate equilibrium temperatures based on experimental ionic liquid inhibitor data.
• Screening potential ionic liquid inhibitors for gas hydrate formation based on model predictions.
• Developing real-time decision-support systems for pipeline flow assurance engineering.

Strengths

• Contains 946 experimental data points, providing a basis for model training.
• Systematically evaluates seven distinct feature scaling methods (L1-Normalization, Robust Scaler, etc.), with performance metrics reported.
• Reports specific model performance results, such as an R² of 0.9156 achieved with Quantile Transformation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.
• The dataset is small at 237.3 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Likely compiled from experimental studies, as the description references 946 experimental data points.

Freshness

Last updated 2026-06-01 20:45:01; freshness should be verified.