Szkoła Główna Handlowa w Warszawie - Centralny System Uwierzytelniania
Strona główna

Time Series Analysis and Forecasting 132011-D
Wykład (WYK) Semestr letni 2023/24

Informacje o zajęciach (wspólne dla wszystkich grup)

Liczba godzin: 30
Limit miejsc: (brak limitu)
Zaliczenie: Ocena
Zakres tematów:

Features of economic time series. Types of forecasts. Forecasts in economics, business and finance. Forecasting methods. Ex post measures of prediction accuracy.

Adaptive forecasting. Moving average as forecast. Exponential smoothing methods: Brown, Holt, Winters, Gardner-McKenzie.

Forecasts from decomposition models. Census II. Additive vs. multiplicative seasonality. Seasonal factors. Deseasonalization. Trend-cycle. Kruskall-Wallis test for seasonality.

Forecasting with multiple regression. Stability tests. Ex ante prediction error. Fan charts. Pseudo-out-of-sample forecasting. Rolling window. Diebold-Mariano test.

Forecasting with growth curves. Deterministic vs. stochastic trend. S-type curves. Innovation diffusion models. Dummy variables in forecasting. Combined forecasts.

Dynamic Factor Models (DFM). Principal Components method. Factor models. Stock-Watson models.

ARMA modeling. Autocorrelation. Properties of correlogram and other sample statistics. Box and Jenkins methodology. Information criteria (Akaike, Schwarz) as tools of the model specification. The Ljung-Box and other diagnostic tests. Optimal ARMA models prediction.

Stationary and non-stationary series. The ADF, KPSS and Phillips-Perron tests. Definition of integrated series. Co-trending. Stochastic trends and random-walk process.

Spurious regression. Definition and tests of cointegration. The Engle-Granger approach. The ECM model.

ARIMA modeling. ARMAX models (models with additional explanatory variables). Forecasting ARMA/ARIMA models.

Long-memory processes. Measures of long memory of the process. Elements of spectral analysis. Methods of fractional integration parameter estimation. Classification of processes based on Hurst exponent and fractional integration parameter.

Deterministic and stochastic seasonality. Dummy variables in seasonality modeling. Seasonal integration tests. Cycles and spectral analysis. Forecasts of seasonal time series - SARIMA models.

VAR models. Multivariate information criteria. VECM and the Johansen tests of cointegration. Long-run economic equilibria and cointegrating relationship. Forecasts from the VAR models. Impulse response analysis.

Forecasts from the VAR models. Impulse response analysis.

Nonlinear varying variance models. Introduction to ARCH and GARCH models.

Grupy zajęciowe

zobacz na planie zajęć

Grupa Termin(y) Prowadzący Akcje
1 każdy piątek, 13:30 - 15:10, sala 3D komp
Ewa Syczewska szczegóły
Wszystkie zajęcia odbywają się w budynku:
budynek C
Opisy przedmiotów w USOS i USOSweb są chronione prawem autorskim.
Właścicielem praw autorskich jest Szkoła Główna Handlowa w Warszawie.
al. Niepodległości 162
02-554 Warszawa
tel: +48 22 564 60 00 http://www.sgh.waw.pl/
kontakt deklaracja dostępności mapa serwisu USOSweb 7.0.4.0