Machine Learning for Time Series: Introduction

About

Time is your most valuable feature. Learn to analyze, forecast, and detect patterns in temporal data using both classical statistics and modern machine learning. From ARIMA to XGBoost, from anomaly detection to change points: build the skills that make data speak.

If you would like to do a follow up on this topic, please also check out Machine Learning for Time Series: Advanced

Course leader

Dr. Yuliya Shapovalova and Dr. Roel Bouman

Target group

Bachelor, Advanced Bachelor, Master, PHD, Professional.

Participants are expected to have:

- Minimal programming proficiency in Python, including experience with data manipulation libraries such as Pandas and NumPy
- Basic knowledge of statistics, including concepts like mean, variance, distributions, and hypothesis testing.
- Familiarity with basics of machine learning, such as the difference between supervised and unsupervised learning, training/test splits, and model evaluation metrics.

No prior experience with time series analysis is required; this course starts from the fundamentals. Participants should bring a laptop with a working Python environment (setup instructions will be provided before the course)

Course aim

1. Understand Time Series Fundamentals.
2. Preprocess Data and Perform Feature Engineering.
3. Train and Evaluate Forecasting Models (Classical and ML).
4. Understand Anomaly Detection and Change Point Detection.
5. Apply Methods to Real Data and Avoid Common Pitfalls

Fee info