🎓 🔗 Supercomputing Academy: Time Series Forecasting

Time series forecasting has gained particular attention in recent years because it can provide a fundamental insight into the future by analyzing historical data, assuming that future trends are similar to those that occurred in the past. The primary value of using time series for forecasting is that at business time, future trends and outcomes are not fully available and can only be estimated using historical forecasts. This technology is vital in many fields to understand current seasonal trends and fit their data to the flow of time. Time series forecasting has important applications in weather forecasting, business, healthcare, and finance.

Flexible online course: Combination of self-study and live seminars (HLRS Supercomputing Academy)
Organizer: HLRS, University of Stuttgart, Germany

Prerequisites

• Good experience in Python programming.
• Basic knowledge in machine learning.
• Basic knowledge in Linux.

Content levels

Community-target and domain-specific content: 30 hours

Learn more about course curricula and content levels.

This course is intended for, but is not limited to, the following groups:

• Postgraduate, non-computer scientists (e.g. engineers).
• Researchers, financial and business sectors interested in time series analysis and market trends.
• Everyone who is interested in Time Series Forecasting technology.

Learning outcomes

After this course, participants will:

• learn the fundamentals of time series forecasting analysis.
• gain knowledge of different algorithms and methods in time series forecasting.
• gain basic skills in data preprocessing, data cleaning and preparation.
• learn how to build a variety of time series forecasting models and optimize hyperparameters. In addition to interpreting and visualizing the results.
• deepen your understanding of different models and algorithms through hands-on exercises and assignments.

Agenda

• Week 1: Introduction into Time Series Forecasting, Long Short-Term Memory (LSTM), Exponential Smoothing.
• Week 2: Autoregressive integrated moving average (ARIMA), TBATS, Multivariate Time Series Forecasting, DeepAR.
• Week 3: XGboost, N_BEATS, Prophet.

Access to learning blocks will be unlocked one at a time on a weekly basis. The online seminars cover the course topics in the order listed above.

Registration information

Register here.
We encourage you to register to the waiting list if the course is full. Places might become available.

Registration closes on October 12, 2025.