Machine Learning for Time Series: Advanced

About

Take your time series skills further. This advanced course explores powerful methods including deep learning architectures for forecasting, probabilistic approaches, and more complex anomaly detection techniques. Ideal for those with foundational knowledge who want to tackle complex real-world challenges

Course leader

Dr. Yuliya Shapovalova and Dr. Roel Bouman

Target group

Master, PHD, Postdoc, Professional.

Participants are expected to have:

- Completion of Machine Learning for Time Series: Introduction or equivalent knowledge, including familiarity with time series fundamentals (stationarity, autocorrelation), ARIMA models, and basic anomaly detection.
- Solid programming proficiency in Python, including experience with Pandas, NumPy, and scikit-learn.
- Good understanding of machine learning concepts, including model training, evaluation, and overfitting.
- Basic familiarity with neural networks (e.g., understanding of layers, activation functions, and backpropagation) is helpful but not required.
- Participants should bring a laptop with a working Python environment (setup instructions will be provided before the course)

Course aim

1. Understand Deep Learning Architectures for Time Series.
2. Apply Probabilistic Approaches and Uncertainty Quantification.
3. Implement Advanced Anomaly Detection Methods.
4. Combine Models Using Ensemble Methods.
5. Apply Advanced Methods to Complex Real-World Problems

Fee info