Subscribe Bookmark

Community Manager


Feb 27, 2013

Building forecasting models for time series data

Time series data have a natural ordering by time, and that order is relevant to the analysis of the data characteristics and the way they are used to forecast future events.

On Oct. 27, Robert Anderson, Systems Engineer based in the JMP UK office, presented a webcast on building forecasting models for time series data. By popular request, his demo is now available, on demand, in two episodes.

In the first episode, Robert explores seasonal time series data visually and then builds a variety of time series models that he assesses based on their AIC and MAPE scores. He uses the Winters method, seasonal exponential smoothing, ARIMA and linear regression to build the models. In the second episode, he uses data held back from the original data set to evaluate the models and determine their forecasting potential.

In the JMP sample data, you will find the Seriesg data Robert uses. (HELP>SAMPLE DATA)

Our Mastering JMP webcasts on Thursdays have concluded for 2011. Please check back in January for a list of live webcasts that delve into ways to apply JMP or JMP Pro to the challenges you face in your work.