Forecasting and Time Series Analysis

Course Description
Of fundamental importance in business planning is the ability to generate a usable forecast. Although time series is most often applied to economic sequences of events, the methodology can be applied to any manufacturing, engineering, or research data that are collected over time with a stable input pattern. This course will provide participants with a working knowledge of the applications of forecasting and time series analysis along with a general understanding of the underlying theory. This course combines both theory and application and focuses on the analysis and modeling of a series of events with the object of generating meaningful forecasts.

Who Should Attend
This course is designed for business managers, engineers, and researchers involved in modeling processes that occur in time, such as economic and business series, manufacturing quality, yield, and various types of naturally occurring phenomena (such as temperature or rainfall, etc.).

Learning Objectives
Through training, participants will:

  • Know how to characterize trends, seasonal components, and autocorrelation
  • Be able to use a variety of techniques for forecasting applications including simple and multiple regression; moving average models; single and double exponential smoothing models; and ARIMA models.
  • Understand how to use time series models for forecasting and how to analyze forecast errors with statistical software.

Course Outline
Introduction

  • Terminology
  • Types of Forecasting Problems
  • Demand Patterns and Decomposition
  • Analysis Roadmap
  • Forecasting Accuracy and Level
  • Tracking Signals
  • Importance of Forecast Error
    • Application of Forecast Error to Inventory

Univariate Forecasting Techniques

  • Moving Average Smoothing
  • Single Exponential Smoothing
  • Double Exponential Smoothing
  • Winter’s Seasonal Model
  • Autocorrelation Function
  • Partial Autocorrelation Function
  • ARIMA Models

Multiple Variable Forecasting Techniques

  • Handling Dates
  • A Univariate Regression Example:  When will the 100m world record be broken?
  • Selecting Variables for Multivariate or Causal Forecasting
    • Cross Correlation
  • Testing First Order Autocorrelation
  • Multiple Regression Approach

Prerequisites
Basic SPC or the equivalent

Course Format
8 hours
Instructor-led class training, with opportunities to practice learned skills using prepared data
Minitab or JMP Statistical Software

Follow QSG on LinkedIn!
Become a QSG Member today!

Always Keep Improving!