Diese Anwendung erfordert Java-Skript.

Bitte aktivieren Sie Java-Script in den Browser-Einstellungen.
Lecture Advanced Time Series Analysis, 14. Lesson
Grammig, Joachim (2020)
clipboard
mla
clipboard
Grammig J. "Lecture Advanced Time Series Analysis, 14. Lesson.", timms video, Universität Tübingen (2020): https://timms.uni-tuebingen.de:443/tp/UT_20201123_002_ws2021atsa_0001. Accessed 12 Dec 2024.
apa
clipboard
Grammig, J. (2020). Lecture Advanced Time Series Analysis, 14. Lesson. timms video: Universität Tübingen. Retrieved December 12, 2024 from the World Wide Web https://timms.uni-tuebingen.de:443/tp/UT_20201123_002_ws2021atsa_0001
harvard
clipboard
Grammig, J. (2020). Lecture Advanced Time Series Analysis, 14. Lesson [Online video]. 23 November. Available at: https://timms.uni-tuebingen.de:443/tp/UT_20201123_002_ws2021atsa_0001 (Accessed: 12 December 2024).
file download bibtex   endnote
Information
title: Lecture Advanced Time Series Analysis, 14. Lesson
alt. title:
creator: Grammig, Joachim (author)
subjects: Wirtschaftswissenschaft, Advanced Time Series Analysis, Lecture
description: Vorlesung im WiSe 2020-2021; Montag, 23. November 2020
abstract: Students master state-of-the-art time series econometrics, both univariate and multivariate. They apply time series methods with awareness of their potential and limitations in macroeconomics and finance. They command an econometric programming language independently and productively to perform empirical analyses involving time series data. They present and discuss their results of the application of time series methods in a scientific fashion. The module deals with a rigorous treatment of state-of-the art univariate and multivariate time series methods used in economics and finance. This includes: 1. Autoregressive moving average models 2. Forecasting 3. Regression analysis with stationary and non-stationary time series 4. Unit root tests 5. Structural vector-autoregressive models and cointegration 6. Equilibrium correction and Johansen methodology 7. Amplification of time series methods in macroeconomics and finance using econometric software 8. Conditional heteroskedasticity in financial time series
publisher: ZDV Universität Tübingen
contributor: ZDV Universität Tübingen (producer)
creation date: 2020-11-23
dc type: image
localtype: video
identifier: UT_20201123_002_ws2021atsa_0001
language: eng
rights: Url: https://timmsstatic.uni-tuebingen.de/jtimms/TimmsDisclaimer.html?638695930330869777