"Unobserved Component Models (UCM) - time series forecasting"

dieffendieffen Member Posts: 2 Contributor I
edited June 2019 in Help
I am working on a project to forecast the arrival of patients to the emergency room within a specified time (future) period.  Still consider  myself a novice both in Rapidminer and data science but have a passion to learn.  I ran into a technique called Unobserved Component Models (UCM) that seemed to handle (decompose) seasonalities, cycles, trends, etc.  Intuitively there are "seasonal" factors like the day of the year, weekday / weekend, weather conditions, holiday / non holiday, etc.. that would be a factor on ER demand.  When I searched in the operators in Rapidminer I didn't see UCM available.  I didn't know if anyone had some suggestions, atlernatives, etc.  Availalbe in R or Weka extension?

Appreciate any help.

Best regards,
Eric
Tagged:
Sign In or Register to comment.