Introduction to active learning for classification

Active learning is, in essence, about maximizing the performance gain (of an underlying, supervised machine learning algorithm) per label. Some possible (query) strategies and the core problem emerging, being sampling bias, are discussed in the introduction to active learning. Then an effective way of circumventing it by balancing exploring and exploiting accordingly is given via Thompson sampling. Lastly active learning is put to use for time series classification. Concepts are outlined in a heuristic manner.

Speaker at introduction to active learning: Simon Bachhuber is currently writing his master’s thesis in physics at the University of Regensburg. The topic is active learning and its practical implementations. The ulterior motive is to (ultimately) implement active learning in collaboration with Syskron to improve label efficiency.

This meeting is hosted by the Bavarian Industry Association for Applied Artificial Intelligence e.V.  The purpose of the association is the cross-sectoral exchange of knowledge between Bavarian industry by promoting science, research and popular education. In concrete terms, this is the promotion of the dissemination of knowledge about processes and procedures of applied practical intelligence (AI) and machine behavior (ML). This should include, in general, securing the right to competitiveness of industrial companies and maintaining that the Free State of Bavaria can position itself as the leading region for applied AI technology and processes.

Due to the current situation of Covid-19, all our meetings are carried out online. So if you are interested in attending, you just have to be online. For more informatio have a look at the link below.

Thursday 2020-06-18, 18:00, online event

Information for participation & more details:

Last edited on