Fundamentals of Forecasting in SAS
Wednesday, May 27
Course Description:
The SAS system has many powerful tools to leverage time series data to forecast future values and outcomes. This class provides a solid grounding in the fundamentals for forecasting with SAS. The course begins with handling time series data, including lagged variables and seasonality. Forecasting methods include time series regression, ARIMA, and ARCH / GARCH. New this year are Unobserved Components and working with an interrupted time series, where a sudden change in the history of the data impacts future forecasting. The interrupted time series will be especially useful for people working with federal data sources affected by recent changes in collection and content.
Course Outline (each part in 1 hour):
Part 1: Time Series Data Management
Part 2: Forecast Factors
Part 3: ARIMA Models
Part 4: Advanced Forecasting Methods
The SAS system has many powerful tools to leverage time series data to forecast future values and outcomes. This class provides a solid grounding in the fundamentals for forecasting with SAS. The course begins with handling time series data, including lagged variables and seasonality. Forecasting methods include time series regression, ARIMA, and ARCH / GARCH. New this year are Unobserved Components and working with an interrupted time series, where a sudden change in the history of the data impacts future forecasting. The interrupted time series will be especially useful for people working with federal data sources affected by recent changes in collection and content.
Course Outline (each part in 1 hour):
Part 1: Time Series Data Management
- Time Series Data
- Imputation of Missing Data
- PROC TIMEDATA
Part 2: Forecast Factors
- Lagged Data
- Seasonality
- Time Series Regresion
Part 3: ARIMA Models
- ARIMA - Moving Average
- ARIMA - Autoregressive
- Bayesian ARIMA using PROC MCMC
Part 4: Advanced Forecasting Methods
- ARCH / GARCH Models
- Interrupted Time Series Forecasting
- Unobserved Components
|
About the Instructor
David J Corliss, PhD has been working with SAS since 12693. He is a Principal Data Scientist at Grafham Analytics, a data science consultancy and is a frequent presenter and instructor at SAS conferences with expertise in time series data and methods, and ethical best practices in data, analytics, and AI. Dr. Corliss is the founder of Peace-Work (link: www.peace-work.org), a volunteer cooperative of statisticians, data scientists and other researchers applying analytics in issue-driven advocacy.
|
