Training Module: Time Series Analysis and Forecasting


This time series analysis training module covers the use of STATGRAPHICS in analyzing time series data and for applying statistical forecasting methods.

Module length: 0.5 day

Outline:

Descriptive Methods

  • Time Sequence Plots
  • Subseries Plots
  • Autocorrelations and Partial Autocorrelations
  • Periodograms
  • Tests for Randomness
  • Cross-correlation Functions

Smoothing

  • Weighted Moving Averages
  • Nonlinear Resistant Smoothers
  • Combining More than One Smoother

Seasonal Decomposition

  • Time Series Models
  • Estimating the Components
  • Generating Seasonally Adjusted Data

Forecasting

  • Moving Averages
  • Trend Models
  • Exponential Smoothing
  • ARIMA Models
  • Selecting a Model For Statistical Forecasting Methods
  • Automatic Model Selection