An autoregressive integrated moving average (ARIMA) model can be used to predict (forecast) future points in a time-series while automatically detecting and correcting for anomalies, seasonality, and holiday effects. It is a popular tool for data scientists to estimate future demand, for example when creating retail sales or manufacturing production forecasts. Using this Block, Looker developers can add this advanced analytical capability right into new or existing Explores, no data scientists required.
Using this Block, you can integrate Looker with BigQuery ML Time-series (ARIMA Plus) models to get the benefit of forecasting with advanced analytics without needing to be an expert in data science. BigQuery ML ARIMA Plus model includes the following functionality:
- Infer the data frequency of the time series
- Handle irregular time intervals
- Handle duplicate timestamps by taking the mean value
- Interpolate missing data using local linear interpolation
- Detect and clean spike and dip outliers
- Detect and adjust abrupt step (level) changes
- Detect and adjust holiday effects
- Detect and adjust for seasonal patterns
This Block gives business users the ability to do time-series forecasting from a new or existing Explore. Explores created with this Block can be used to train multiple time-series models, evaluate them, and access their forecasts in dashboards or custom analyses.
Learn more in the associated BigQuery ML Tutorial.
This Block can be installed via the Looker Marketplace.