Beam - Overview
Find out how to use Beam for correlation and decomposition
What is Beam
PredictHQ's Beam helps you identify what events have impacted your demand in the past, to help you make better decisions in the future.
Beam is an automated correlation engine that accurately reveals the events that drive demand for your business. Beam converts unstructured, dynamic event data into a workable dataset, allowing data science teams to use intelligent event data in machine learning models. PredictHQ Beam combines and correlates our event data with your transactional demand data. See our Beam Product Page for more details.
Beam is accessible via our Control Center Web Application and via the Beam API. To use Beam you upload demand data for a location. Beam decomposes your data into a baseline and remainder and shows you the correlation with events data. Beam will also show you the most important categories that are impacting your location. You can also export the decomposed data from Beam.
The Beam UI is good to use if you have a smaller number of analyses or if you are exploring how events impact your demand. Use the Beam API if you want to bulk upload a large number of analyses. Many customers use Beam to analyse which categories impact them the most and then use that information as input to building machine learning features.
Beam is free to use and available to all customers with Control Center access. Simply sign up for a free account and upload your data to Beam.
See below for an introduction to Beam
How to use Beam
See the following articles for details on how to use Beam:
The free version of Beam in Control Center allows you to perform a correlation analysis for individual locations. Please get in touch with us if the following is of interest:
If you want to use Beam for a large number of locations please contact us.
The PredictHQ data science team can perform a more detailed analysis of your demand data to show you what categories are most impacting your demand.
The PredictHQ data science team can help you to build event-based machine learning features for use in your demand forecasting.