Seminar - Introduction to time series in R
Introduction to time series in R (Prüfungsnummer 3061)
(Selected Topics in Quantitative methods, in English)
The meeting will take place on:
|Montag, 18.10.2021||13 - 16 Uhr||HG 217||I. Semeniuk / K. Gaykalov|
|Montag, 25.10.2021||13 - 16 Uhr||HG 217||I. Semeniuk / K. Gaykalov|
|Montag, 01.11.2021||13 - 16 Uhr||GD 04||I. Semeniuk / K. Gaykalov|
Time series is one of the most popular topics in the modern Applied Statistics. Different models were developed to describe the behavior of observations of a variable obtained in time. In this course we consider the basic models: ARMA-, GARCH-models, their effectiveness in the description of the time dependent stochastic processes, their implementation in R. The basic knowledge of these models will let a student study other existing time series models faster, understand the principles of their construction (parameter estimation, components selection) and apply them to the real data (goodness of the model, forecast). R programming language allows to make data analysis easy, fast, fun, beautiful and creative. The use of R becomes more popular because of many reasons. That is why the time series analysis in the class will be performed in R.
Next topics will be considered in the course:
- Examples of Time Series. Time Series Plots in History. Time Series and Stochastic Processes. Means, Variances and Covariances. Stationarity.
- Models for stationary time series. Auto Regressive Moving Average Processes. ARIMA models. Backshift operator. Parameter estimation. Diagnostics and forecast in R.
- GARCH models. Parameter estimation. Diagnostics and forecast in R.
- Seasonal models.
- Time series regression models.