Miami Housing Prices for Regression Model Evaluation, 2016

A 2016 dataset of Miami, Florida housing information, curated for the TabArena Tabular ML IID Study. It is intended for evaluating predictive machine learning models on independent and identically distributed tabular data, with a regression task. The original data was sourced from a 2021 academic paper by Bourassa et al. on urban house prices.

Use Cases

• Benchmarking regression model performance based on the dataset's stated focus on IID tabular data.
• Predicting housing prices based on the dataset's intended regression task.
• Analyzing urban real estate factors in Miami based on the dataset's geographic and topical focus.

Strengths

• Curated for a specific machine learning study (TabArena Tabular ML IID Study), suggesting a clear evaluation purpose.
• Duplicated homes were removed during curation, which likely improves data integrity.
• Has a clear license (CC BY-NC-SA 4.0) and a cited academic reference for provenance.

Limitations

• Column names, sample data, file formats, and exact row count are unknown, limiting pre-download assessment.
• An anomaly exists where the dataset has approximately 13k samples, while the original paper describes 57k.
• The last update date and original organization are unknown.

Provenance

Source

Original data from OpenML (ID 43093), based on research by Bourassa et al. (2021).

Collection Method

Curated by the TabArena team, which dropped duplicates and the ID column.

Time Range

2016

Freshness

Dataset year is 2016; last update date is unknown.

Geography

Miami, Florida, USA