Free Machine Learning Sample Data Sets

Download data sets to hone your skills in machine learning. All files are .csv format.

Register for a free Squark account and see the power of automated machine learning for actionable predictions.

Here are sample Machine Learning datasets for use with Squark.  These datasets support Squark’s supervised machine learning capabilities for “how much?” predictions such as forecasting (regression), as well as classification predictions “which of two things?” (binary), or “which of three or more things?” (multivariate or multinomial).

The training files contain historic data with a target label for learning. The target column (also know as the dependent variable) is identified; however, in many datasets you may be able to use other columns as targets. The production datasets contain new data on which you want to make predictions. For more, view this What Are Training and Production Data video. If you have any questions, please contact us at info@squarkai.com.

Use this customer data from a telecom to predict which customers will stop being customers (churn) and those that will remain customers. The target variable in the training file is “Churn.”

Use this customer data to predict which offers to send to new customers.  The target variable in the training value is “Contract.”

Use this customer data to predict the price to offer new customers for new services. The target variable in the training data is “Price.”

Use this scientific data about flowers to predict which each flower’s species. The target variable in the training data is “species.”

Use this Amazon Alexa data to predict which products will receive customer feedback. The target variable in the training data is “feedback.”

Use this Amazon Alexa data to predict which products will receive customer feedback. The target variable in the training data is “feedback.”

Use this customer data to identify which offers, in the form of coupons, to send to customers. The target variable in the training data is “coupon_id”.

Use this banking data about customer lending profiles to predict lending default. The target variable in the training data is “is_bad.”

Use this online advertising data to predict which Google ads will generate more than one transaction.  The target variable in the training data is “Purchase >1 Time.”

Use this choice data from automobile survey research to identify which combinations of features to offer car shoppers.  The target variable in the training file is “choice.”

Use this media mix data to predict sales.  The target variable in the training data is sales.

Use this Facebook online advertising data to predict the numbers of interactions each ad unit will generate.  The target variable in the training file is the “total interactions.”

Use this Facebook online advertising data to predict which ads should based be placed on their predicted lifetime value segment. The target variable in the training data is “type.”

This automotive marketing dataset enables predicting lifetime value. Use the target variable “Customer Lifetime Value” in the training file dataset

This product data contains information about wine ingredients so the products can be segmented by customer preference.  Use the target variable “customer segment” in the training file to predict segments.

This tag auditing and monitoring data identifies information about the marketing tags embedded on a website.  Use the target variable ” Tag Load Time” in the training file to predict how long it will take tags to load.

This sales data contains information about customers and their transactions. Use the target variable “Deal Size” in the training file to predict the opportunity size category.

This customer data contains information about behaviors and transactions. Use the target variable “target” in the training file to predict if the customer should be targeted with marketing communications or not.

This customer data contains information about their bank account and their customer type.  Use the target variable “Res_Type” in the training file to predict their response type.

Subscribe here for alerts on new data sets, plus AI and machine learning info from Squark.

Change subscriptions any time.