BigQuery has recently introduced few GenAI functions using which you can forecast, evaluate your forecast and even detect anomalies from within a simple SQL query itself. These are the AI.FORECAST, AI.EVALUATE and AI.DETECT_ANOMALIES functions. These obviously work on time series data, and you can start playing around with them with any of the public datasets on BigQuery. I have used the GA4 ecommerce public dataset to try these out. I discuss the underlying theory (the TimesFM model) as well as the queries and how to visualize them and finally a comparison with other bespoke techniques for time series forecasting and anomaly detection using Python packages.
Here is the document shown in the video:
https://docs.google.com/document/d/1P_NwHKSDMMsCOGM_RIVUn3Jk8Qo5e4I0r3v2-yKz7VQ/edit?tab=t.0
#marketinganalytics #marketingmeasurement #performancemarketing #bigquery #forecasting #predictiveanalytics
Download
0 formats
No download links available.
BigQuery AI Forecast, Evaluate & Anomaly Detection Functions | NatokHD