In this seminar, different approaches for conducting multistep Time
Series Forecasting (TSF) are discussed mainly the following :

• Different Transformer-based Approaches, including different types of attention mechanisms.
• Multi-Layer Perceptron (MLP)-based approaches including different multiresolution approaches.
• Convolution Neural Networks (CNN), being used beyond Computer vision as a time series data encoder.
• Utilizing transformation of time series data into the frequency domain.

The following student study group should be eligible for carrying out the course :

• Für Studierende der Bachelorstudiengänge Wirtschaftsinformatik und Informationsmanagement und Informationstechnologie BSc 4-6
• Master Data Analytics Seminar I (should be carried out before Seminar II and Seminar III).