Our course "Time Series Analysis with Statistical Methods" provides a compact two day introduction into the foundations of statistical learning for forecasting / predictions. The task is the following: given a time series, i.e. a collection of measurements for different time points, one searches the most probable value of this measurement for future time points. The topics include necessary preprocessing steps for numerical data transformations, an introduction into statistical regression methods, neural networks, and support vector machines (SVM), and a discussion of validation methods in order to measure the goodness of the predictions. These methods are especially useful for numerical predictions from series data as they often occur in financial markets but also in production settings and many others. Many practical exercises will be performed with the data mining suite RapidMiner. This ensures that the participants will be able to transfer the gained knowledge to own prediction problems.

Details

• Course ID: 1201
• Number of days: 2 days
• Location: Dortmund, Germany
• Target audience: users, decision makers, analysts, developers, administrators
• Previous knowledge: basic knowledge of computer programs and mathematics
• Methods: lectures, discussions, individual and group work, exercises on realistic data. Participants may introduce own work and project specific questions in order to find particular solutions together with the trainer and other participants.
• Content: this course is a compact introduction into the foundations of statistical learning for predictions and forecasting. It addresses beginners and intermediate learners. Topics of this course are
  • Definition of prediction / forecasting of series data
  • Statistical Learning: Regression Methods, Neural Networks, Support Vector Machines (SVM)
  • Preprocessing for series data: from series to data points
  • Forecasting with regression methods
  • Influence of the prediction horizon
  • Validation of forecasting: introduction into performance criteria, cross validation, bootstrapping
  • Visualization of series data and predictions: high-dimensional data visualizations
  Extensive exercises on different data sets will be performed for all topics.

Prices

Value added tax (VAT) may have to be added to these prices.