Skill Introduction

In this course we focus on ARIMA forecasting models. We will combine the concepts of autoregression (the “AR” in ARIMA) and moving average (the “MA” in “ARIMA”) for time-series data with the concept of integrated (the “I” in “ARIMA”). Integrated refers to the differencing that occurs for a time series to be stationary. 

Stationarity is the first and necessary step in analyzing time-series data with ARIMA models. You will learn how to identify if a time series is stationary or not and how to make nonstationary data become stationary by differencing using functions in R. Once data is stationary, you can start exploring various ARIMA models.

You will then be introduced to a process for exploring combinations of parameters that are likely to optimize ARIMA model forecasts. You will also learn how to use a function in R to automatically create an ARIMA model. Finally, you will critique various ARIMA models based on plots and performance metrics.



Course curriculum

  • 1

    Orientation

    • Introduction Video: Jose Rodriguez - ARIMA in Practice

    • Introduction to the Skill

    • Glossary

  • 2

    Content and Activities

    • Stationarity Introduction

    • Stationarity Differencing

    • Knowledge Check 1

    • ARIMA Introduction

    • ARIMA Components

    • Knowledge Check 2

    • ARIMA Model and R Example - Part 1

    • ARIMA Model and R Example - Part 2

    • Knowledge Check 3

    • ARIMA Model and R Example - Part 3

    • ARIMA Model and R Example - Part 4

    • ARIMA Model and R Example - Part 5

    • Knowledge Check 4

  • 3

    Application Exercise

    • Instructions

    • Exercise Files

    • Debriefing

  • 4

    Summary

    • Conclusion

    • Final Quiz

    • Survey Instructions

    • Feedback Survey

    • Survey Verification

    • Next Steps

Learning Outcomes

Upon successful completion, you will be able to:

  • Develop an understanding of stationarity, identifying two forms of it and their importance in time series analysis

  • Perform stationarity testing procedures: conduct a stationarity test, read the ADF test results, make a nonstationary series become stationary, and determine the order of integration

  • Interpret the various components of ARIMA parameters, and identify the procedure of modelling ARIMA forecasts

  • Identify opportunities for utilizing ARIMA forecasting model in practice

Begin your learning today.