Predictive maintenance dataset for demo purposes. This dataset contains fake data and emulates a predictive maintenance use case in a Factory. The objective class is very unbalanced. Anomaly detectors isolate well the failures
Madrid air pollution data, to predict pollution alerts 1 day in advance. Each row represents a day in central Madrid including features about: NO2 air levels, weather, traffic information with the objective field pollution "YES/NO". The objective field has been defined using Madrid local government pollution alerts definition. Contains data from 2013 to 2017.
Issued tickets for every sale between May and August of 2018. Columns: FolioUnique id for every ticket, HoraDatetime of the sale, TotalTicket amount in Mexican Pesos, CajeroPartial name of the cashier.
Source: https://www.kaggle.com/agasca/retail-sales/version/1
Predictive maintenance dataset about oil wells downhole equipment failures using sensor data.
Source including more information: https://www.kaggle.com/c/equipfailstest/overview
Actual source accessible data: https://raw.githubusercontent.com/geooot/tamudatathon2019/master/equip_failures_training_set.csv
Missing values have been either removed or replaced with the mean as a cleaning step providing good anomaly detectors results.
Madrid air pollution data, to predict pollution alerts 4 days in advance. Each row represents a day in central Madrid including features about: NO2 air levels, weather, traffic information with the objective field pollution "YES/NO". The objective field has been defined using Madrid local government pollution alerts definition. Contains data from 2013 to 2017.