Retail fraud detection dataset (labeled)

POS transactions with a labeled anomaly column marking suspicious activity (odd-hour high-value bulk buys) — a starter for retail fraud and anomaly detection.

Retail POSSeeded - reproducibleCSV + Excel100% in-browser

Generate & download

Save / load scenario (stored only in this browser)

Good for

Fraud detectionAnomaly detectionImbalanced classificationRule vs. model comparison

Why this dataset is realistic

The catalog is organized into real affinity groups (e.g. chips + salsa + soda) that co-occur within baskets, so an association-rule miner actually surfaces lift — exactly what a market-basket exercise needs.

Need to change the size, seed, or columns? Open the full Retail POS generator to customize and re-export. Want the same data as everyone else? This page uses a fixed seed, so the download is identical every time.

Related sample datasets

FAQ

Is the data real?

No - it's 100% synthetic and generated in your browser. It contains no real people or companies and is free to use commercially.

Will I get the same file each time?

Yes. This page fixes the seed, so the dataset is reproducible. Clear the seed in the generator above for fresh random data.