Hi Community!
I was wondering if there was a way to integrate the power of Predictive tools with Time Series forecasting.
Namely, I'd like to be able to use historical data to forecast revenue for the next 12 months, but to be able to do so while taking into account specific predictor variables.
As an example, revenue has been heavily dependent on COVID cases, which in turn are dependent on vaccination rates (at a statistically significant level.)
Is there any code-free way to build something like this out? Otherwise, still open to suggestions on coding something up as well 🙂
Thanks as always!
Solved! Go to Solution.
Hi
Alteryx has several built-in Time Series tools, but they use specific algorithms ARIMIA and ETS. What you are describing, with independent and dependant variables sounds more like a Regression-based model.
Here is a good overview of all the Predictive tools within Alteryx
https://help.alteryx.com/20213/designer/predictive-analytics
Here is a great article that will let you use the results from a Regression model to generate your forecast.
https://statisticsbyjim.com/regression/predictions-regression/
Hope this helps!
Hello @AleksM ,
In complement of @Laurap1228 you can focus on the ARIMA algorithm. One of the difference with the ETS is that you can specify some covariates, for instance Covid cases to forecast your Revenue. Just activate the option (as seen below). And the prediction will be based on both your revenue historic but also on the covariate historic.Youjust need to blend the data in the same table before use this algo.
Some additional detailed: How to Arima tool
And of course the Alteryx DataScience learning path that have a dedicated focus on the Dynamic Time Series: Data Science Learning path
Hope it helps