Training Dates - please open the booking form by clicking the listed date.

training duration icon Statistics - Time Series Analysis

Duration: 2 Days
Delivery Type: Classroom
Target Audience: Data Analysts
Course Number: 2020631
Method: Lecture with examples and exercises.
  • Topics
  • Summary
  • Trainer
  • Download
  • Conditions

Tab 1


A. Univariate analysis of time series data

[Duration: 0.25 Days] Estimation of the moment-generating functions (expected value, auto-covariance) - auto-correlation: the lag operator, creating and interpretating the correlogram - smoothing of time series data: moving averages, exponential smoothing - transformation and filtering of time series data - first-order and second-order differences

B. Decomposition of time series using deterministic models

[Duration: 0.5 Days] Component models: additive and multiplicative models - seasonal structures in time series: trend, seasons and identification of the seasonal pattern, prognosis and residual analysis - level shifts - linear, parabolic, logistic, exponential fit and regression of time series - polynomials - quality measures

C. Periodicities in time series

[Duration: 0.25 Days] Trigonometric functions and their importance for periodic trends - period detection and frequencies - periodogram: identification and interpretation - regression models with periodic oscillations - spectra and spectral density estimation of time series - introduction to Fourier transformation for time series

D. Univariate linear time series models using AR(I)MA

[Duration: 0.25 Days] Stationarity in time series - White Noise process - AR (Auto Regressive)-models - MA (Moving Average)-models - ARMA and ARIMA models - forecasting - residual analysis - statistical tests for linear time series models - quality measures and model selection

E. Analysis of multidimensional time series

[Duration: 0.25 Days] Cross-correlation and cross-covariance - stationary cross-covariance - co-integration - introduction to cross-spectral analysis and coherence analysis

F. Multidimensional time series using VAR

[Duration: 0.25 Days] VAR (Vector AutoRegressive) processes: modeling, prediction, residual analysis, quality measures, tests

G. Time series with exogenous influences

[Duration: 0.25 Days] Regression with auto-correlated shocks - intervention analysis - transfer function models

Tab 3

Tab 4

Tab 5

Tab 6

For trainings with hands-on labs you are kindly requested to bring your own laptop with the required software. Alternatively, you can use our virtual machines (VMWare) for the MS Windows OS.
These course dates don´t fit in your calendar?
We also offer individual appointments or on-site trainings for your team and charge a per-day fee depending on the size of the group.
In order to enroll for a training or for general inquiries regarding our training and consulting services please use the contact form, send us an email or phone. Kontaktformular oder die folgenden Möglichkeiten:
  • Phone : +1.305.395.7962
  • Fax : +1.305.395.7964