E-Commerce Shipping Data for Predictive Model Evaluation

E-CommereShippingData was curated by the TabArena team for their Tabular ML IID Study in 2021. The dataset originates from a Kaggle source and is intended for classification tasks on independent and identically distributed tabular data. Its original purpose was customer analytics in an e-commerce shipping context.

Use Cases

• Predicting shipping outcomes based on e-commerce logistics features mentioned in the description
• Evaluating classification model performance on IID tabular data
• Analyzing potential relationships between warehouse operations and customer delivery results
• Benchmarking data preprocessing methods for datasets with noted anomalies

Strengths

• Dataset is explicitly curated for a published machine learning study (TabArena Tabular ML IID Study)
• License is Public Domain (us-pd), allowing broad use
• Original source and reference are provided for citation and further details

Limitations

• Column-level documentation, sample data, file formats, and row count are unknown
• The curators note potential data issues with the 'Warehouse_block' feature and a disconnect between the target task and features
• Last update date is unknown; freshness unverified

Provenance

Source

Original data sourced from Kaggle: 'Customer Analytics' by Prachi Gopalani (2021)

Collection Method

Curated by the TabArena team; they dropped an ID column and renamed the target feature.

Time Range

Dataset Year: 2021

Freshness

Dataset Year is 2021; last updated date is unknown.

Geography

unknown