Session 1 – 01:20:00 – Introductory concepts in time series analysis.
White noise, stationarity, autocovariance and autocorrelation.
Session 2 – 01:20:00 – Useful plots in time series analysis.
Time plots, seasonal plots, autocorrelation plots. Time series decomposition: additive and multiplicative using the fable package in R.
Session 3 – 02:00:00 – Benchmark forecasting methods.
The naïve, mean, drift, and seasonal naïve methods.
Session 4 – 03:00:00 – Exponential smoothing. Simple exponential smoothing,
Holt’s linear method, Holt-Winters seasonal method, and fable’s general ETS method.
Session 5 – 03:00:00 – Independent practical 1 – Time series plots.
This practical is not compulsory, you can complete this after the course.
Session 6 – 03:00:00 – Exponential smoothing.
Simple exponential smoothing, Holt’s linear method, Holt-Winters seasonal method, and fable’s general ETS method.
Session 7 – 03:00:00 – Independent practical 2 – Time series decomposition and benchmark forecasting methods.
This practical is not compulsory; you can complete this after the.
Session 8 – 02:00:00 – Autoregressive (AR) and moving-average (MA) models.
Unit root tests for stationarity. How to identity the order of an AR(p) or an MA(q) model using autocorrelation and partial autocorrelation plots.
Session 9 – 02:00:00 – Autoregressive integrated moving average (ARIMA) models and seasonal ARIMA models.
Automatic order selection for a (seasonal) ARIMA model using fable. Linear regression with ARIMA errors.
Session 10 – 03:00:00 – Independent practical 3 – Exponential smoothing.
This practical is not compulsory, you can complete this after the course.
Session 11 – 02:00:00 – Generalized Additive Models (GAMs). An introduction to semi-parametric regression using splines. Incorporating trends and seasonal components of a time series using a GAM.
Session 12 – 02:00:00 – An introduction to Bayesian modelling. Implementation of random walks, autoregressive, and moving average models using JAGS.
Session 13 – 03:00:00 – Independent practical 4 – ARIMA models
This practical is not compulsory, you can complete this after the course.
Session 14 – 01:20:00 – Modelling the variance as a time series process.
Autoregressive conditional heteroskedasticity (ARCH) and generalized ARCH (GARCH) models. Stochastic volatility models.
Session 15 – 01:20:00 – Continuous time models.
Brownian motion and Ornstein-Uhlenbeck processes. Fitting continuous time series models using JAGS.
Session 16 – 01:20:00 – Multivariate time series. Vector autoregression.
Simple examples using JAGS.
Session 17 – 03:00:00 – Independent practical 5 – GAMs and Bayesian models
5.0
