How many rows does each dataset have?

A few hundred, enough to build and debug a pipeline. For training at scale, open the dataset in GoMask Data Factory and generate up to a million rows with the same columns, at the class balance you ask for.

No. It is synthetic data generated to behave like the real thing, so it is safe to share, publish and use in teaching.

Which format should I use in Python?

CSV or Parquet both load with pandas in one line. Excel, JSON, JSONL, SQL, TSV and XML are available too.

Datasets by use case

Classification Datasets for Machine Learning

Each dataset here has a column to predict: is_fraud, churned, attrition, readmitted, default_flag. The features around it are the ones a real model would use, and the sample rows on each page show you the class balance before you download. Load a CSV into pandas or scikit-learn and start with a baseline.

Fraud

Card and payment transactions with an is_fraud or fraud_label column, plus the features fraud models use: amount, merchant category, entry mode, international flag, device and velocity counts. Judge models on precision and recall, not accuracy.

Finance17 cols · 200 rows

Credit Card Fraud Detection

This dataset provides detailed, labeled records of simulated credit card transactions, including transaction amounts, merchant and cardholder information, and fraud indicators. It is ideal for developing and benchmarking machine learning models aimed at detecting fraudulent activity and reducing financial risk in payment systems. The inclusion of transaction context and cardholder demographics supports advanced analytics and feature engineering.

merchant_category	transaction_type	entry_mode
grocery	purchase	chip
coffee_shop	purchase	contactless
electronics	purchase	online

transaction_idcard_numbertransaction_datetime+11 more

Finance19 cols · 200 rows

Financial Transaction Fraud Features

This dataset provides a detailed, feature-rich record of synthetic banking transactions, including transaction metadata, account and merchant information, contextual behavioral features, and fraud labels. It is ideal for developing, training, and benchmarking machine learning models for fraud detection and anomaly analysis in financial services.

transaction_type	merchant_category	fraud_type
purchase	groceries	–
refund	electronics	fake_merchant
purchase	clothing	–

transaction_idaccount_idtransaction_datetime+13 more

Finance19 cols · 200 rows

Payment Fraud Detection Dataset

This dataset contains detailed synthetic payment transaction records, each labeled with ground-truth indicators of fraud. It includes transaction metadata, customer and merchant identifiers, payment methods, device and location context, and fraud reasons, making it ideal for developing and benchmarking machine learning models for payment fraud detection and risk mitigation.

currency	transaction_status	fraud_label
USD	completed	false
GBP	completed	false
INR	failed	true

transaction_idtransaction_datetimeamount+13 more

Insurance20 cols · 200 rows

Medical Insurance Fraud Detection

This dataset contains detailed synthetic records of medical insurance claims, including patient demographics, provider information, claim amounts, service dates, and labeled indicators of fraudulent activity. Designed for machine learning and analytics, it enables robust research and development of fraud detection models in healthcare and insurance. The dataset supports granular analysis of claim patterns, provider behaviors, and patient demographics to identify and prevent fraudulent claims.

claim_status	patient_gender	provider_specialty
approved	female	Family Practice
approved	male	Psychiatry
approved	female	Cardiology

claim_idpatient_idprovider_id+14 more

Churn and attrition

Customers and employees who left, with the columns that hint why: tenure, satisfaction, usage, pay and policy details. Try logistic regression first, then a gradient-boosted model, and compare what each says about the drivers.

Human Resources26 cols · 200 rows

Employee Attrition Prediction Dataset

This dataset provides detailed HR records for employees, including demographics, job history, satisfaction, performance, and attrition status. It is ideal for building predictive models to identify turnover risks, analyze workforce trends, and inform retention strategies. Rich features enable advanced analytics for HR decision-making and organizational planning.

first_name	last_name	marital_status
Riya	Patel	Married
Carlos	Dominguez	Single
Aisha	Okafor	Widowed

employee_idgenderdate_of_birth+20 more

Insurance31 cols · 200 rows

Insurance Policyholder Churn Insights

This dataset provides a comprehensive view of insurance policyholders, their demographic details, policy information, claims history, and churn status for both life and auto insurance products. It is designed to support predictive modeling of customer attrition, enabling insurers to identify at-risk customers and develop targeted retention strategies. The inclusion of satisfaction scores, contact history, and churn reasons makes it ideal for advanced analytics and customer experience optimization.

first_name	last_name	policy_type
Amelia	Richards	auto
Omar	Al-Mansouri	life
Priya	Mehra	auto

policyholder_iddate_of_birthgender+25 more

Gaming16 cols · 200 rows

Game Session Telemetry Dataset

This dataset contains detailed logs of simulated online gaming sessions, including player identifiers, session timings, actions, feature usage, purchases, and outcomes. Designed for gaming analytics, it enables churn prediction, player segmentation, and feature adoption analysis, offering valuable insights for game developers and startups.

player_segment	session_outcome	device_type
new	abandoned	mobile
casual	win	mobile
new	abandoned	console

session_idplayer_idgame_id+10 more

Risk: readmission and default

Hospital admissions with a readmitted flag, and loans with repayment histories and default flags. Both reward careful handling of dates, so nothing from the future leaks into your features.

Healthcare22 cols · 200 rows

Hospital Readmission Prediction

This dataset provides detailed, labeled records of hospital admissions, including patient demographics, diagnoses, procedures, and outcomes related to readmission. It is designed to support predictive modeling for hospital readmissions, enabling health systems to identify risk factors, optimize care transitions, and reduce costs. The comprehensive schema facilitates advanced analytics and machine learning applications in healthcare quality improvement.

readmitted	admission_type	insurance_type
false	elective	private
true	emergency	medicare
false	emergency	medicare

patient_idadmission_idadmission_date+16 more

Finance30 cols · 200 rows

Synthetic Bank Loan Repayment Patterns

This dataset provides detailed, anonymized records of bank loan origination, repayment, default, and prepayment patterns, enriched with borrower demographic and employment information. It is ideal for developing and validating credit scoring models, analyzing risk factors, and exploring repayment behaviors across diverse borrower segments. The data structure supports advanced analytics and machine learning applications in financial risk management.

repayment_status	borrower_marital_status	borrower_employment_status
current	single	employed_full_time
paid_off	single	employed_part_time
current	married	employed_full_time

loan_idborrower_idorigination_date+24 more

Finance33 cols · 200 rows

Finance Peer-to-Peer Lending Data

This dataset provides detailed, transaction-level records from a peer-to-peer lending platform, including loan terms, borrower and investor attributes, credit scores, repayment history, and loan performance indicators. It is ideal for credit risk modeling, investor analytics, and financial platform optimization, supporting both operational and research applications in alternative lending.

loan_status	borrower_employment_status	investor_type
active	employed	individual
completed	retired	institutional
active	student	individual

loan_idborrower_idinvestor_id+27 more

Every dataset in this collection

10 datasets

Finance17 cols · 200 rows

Credit Card Fraud Detection

merchant_category	transaction_type	entry_mode
grocery	purchase	chip
coffee_shop	purchase	contactless
electronics	purchase	online

transaction_idcard_numbertransaction_datetime+11 more

Finance19 cols · 200 rows

Financial Transaction Fraud Features

transaction_type	merchant_category	fraud_type
purchase	groceries	–
refund	electronics	fake_merchant
purchase	clothing	–

transaction_idaccount_idtransaction_datetime+13 more

Finance19 cols · 200 rows

Payment Fraud Detection Dataset

currency	transaction_status	fraud_label
USD	completed	false
GBP	completed	false
INR	failed	true

transaction_idtransaction_datetimeamount+13 more

Insurance20 cols · 200 rows

Medical Insurance Fraud Detection

claim_status	patient_gender	provider_specialty
approved	female	Family Practice
approved	male	Psychiatry
approved	female	Cardiology

claim_idpatient_idprovider_id+14 more

Human Resources26 cols · 200 rows

Employee Attrition Prediction Dataset

first_name	last_name	marital_status
Riya	Patel	Married
Carlos	Dominguez	Single
Aisha	Okafor	Widowed

employee_idgenderdate_of_birth+20 more

Insurance31 cols · 200 rows

Insurance Policyholder Churn Insights

first_name	last_name	policy_type
Amelia	Richards	auto
Omar	Al-Mansouri	life
Priya	Mehra	auto

policyholder_iddate_of_birthgender+25 more

Gaming16 cols · 200 rows

Game Session Telemetry Dataset

player_segment	session_outcome	device_type
new	abandoned	mobile
casual	win	mobile
new	abandoned	console

session_idplayer_idgame_id+10 more

Healthcare22 cols · 200 rows

Hospital Readmission Prediction

readmitted	admission_type	insurance_type
false	elective	private
true	emergency	medicare
false	emergency	medicare

patient_idadmission_idadmission_date+16 more

Finance30 cols · 200 rows

Synthetic Bank Loan Repayment Patterns

repayment_status	borrower_marital_status	borrower_employment_status
current	single	employed_full_time
paid_off	single	employed_part_time
current	married	employed_full_time

loan_idborrower_idorigination_date+24 more

Finance33 cols · 200 rows

Finance Peer-to-Peer Lending Data

loan_status	borrower_employment_status	investor_type
active	employed	individual
completed	retired	institutional
active	student	individual

loan_idborrower_idinvestor_id+27 more

Questions

How many rows does each dataset have?: A few hundred, enough to build and debug a pipeline. For training at scale, open the dataset in GoMask Data Factory and generate up to a million rows with the same columns, at the class balance you ask for.
Is the data real?: No. It is synthetic data generated to behave like the real thing, so it is safe to share, publish and use in teaching.
Which format should I use in Python?: CSV or Parquet both load with pandas in one line. Excel, JSON, JSONL, SQL, TSV and XML are available too.

What should your data show?

No signup. No card.