Time Series and Forecasting (STAT-436)

List of Topics

Introduction: Definitions and Examples.
trend – seasonality – cyclical

Transformation:  Differences method – Seasonal adjustment.

Forecasting: How to forecast future - adequacy of a forecast - regression forecasting against time series forecasting

Some adequacy measures (MAD, MSE, MAPE).

Decomposition and smoothing of times series: moving averages - exponential smoothing double exponential smoothing.

Stationary Time Series Models: Auto-Regressive processes (AR(1), AR(2), AR(p)), Moving Average processes (MA(1), MA(2), MA(q)), The mixed Autoregressive-Moving Average Model ARMA(p,q).

Forecasting: Minimum Mean Square Error Forecasts for ARMA and ARIMA models.

Forecasting, prediction limits and updating forecasts.

ARIMA(p,d,q) models:  Autocorrelation and partial autocorrelation functions - identification of appropriate model

Fitting models to real and simulated data sets. Diagnostic checks on the residuals.

Case studies: training on how to analyze real life data sets using the statistical package MINITAB - write reports.