The course is offered as “20-00-1135-vl Kontinuierliches Maschinelles Lernen” at TU Darmstadt in person in the summer semester 2023. It is the second iteration of and largely analogous to the previously offered “Open World Lifelong Learning – A Continual Machine Learning Course” in 2022, powered by ContinualAI.org & hessian.AI live-streamed on Youtube.
New updated slides will appear weekly. Recorded lectures are available on last year’s course webpage & on YouTube.
Course Objectives
Machine learning studies the design of models and training algorithms in order to learn how to solve tasks from data. Whereas historically machine learning has concentrated primarily on static predefined training datasets and respective test scenarios, recent advances also take into account the fact that the world is constantly evolving. In this course, we will go beyond the train-validate-test phase and explore modern approaches to machines that can learn continually. In addition to a comprehensive overview of the breath of factors to consider in continual learning, the course will delve into techniques that span mitigation of forgetting across multiple tasks, selection of new data in continuous training, dynamic model architectures, and robustness with respect to unexpected data inputs.
Course Content
- Introduction and Course Overview – Download pdf
- Knowledge Transfer, Transfer Learning – Download pdf
- Knowledge Retention Part 1: Regularization – Download pdf
- Knowledge Retention Part 2: Rehearsal – Download pdf
- Active Learning – Download pdf