Time series analysis and forecasting

In analyzing time series one searches for structures and patterns to describe and explain the underlying process. But also for ways to use adequate models fitted to predict future values or to study the effects of alternative scenarios.

Time series occur in a wide range of disciplines, from business, economic and social sciences to biomedical and engineering contexts. This course  treats actual methods for time series analysis, modelling and forecasting.

Apart from the traditional methods for trend and seasonal decomposition of time series, more advanced statistical techniques, both in the time-domain and in the frequency domain are discussed and underlying principles are explained.

Insight in and practice with time series

In this course:

• You gain insight in current approaches for time series analysis, modelling and forecasting, specifically:
  • Exponential Smoothing models (Simple, Holt, Holt-Winter)
  • Box-Jenkins models (ARMA, ARIMA, SARIMA)
  • Multivariate time series modesl for correlated series (dynamic regression and VARMA models)
• You learn to analyze, to model and to validate time series data using relevant statistical software such as R, Minitab or JMP
• You learn to use the models obtained for time series analysis forecasting and scenario analysis

Intended for

Academics and professionals who have to analyze and predict time series data in their work. The course is also suited for lecturers at universities or colleges of higher education who want to be informed on actual methods for time series analysis.

Knowledge of basic statistical techniques like testing, estimating and regression modelling is assumed.

In consultation with the participants this course can be taught in Dutch or English. Practical exercises are in R.