Daniel Kottke receives prize for dissertation on artificial intelligence

The dissertation "A holistic, decision-theoretic framework for pool-based active learning'' describes basic procedures for the development of an Artificial Intelligence (AI), when information with high costs (for example time expenditure) has to be requested from experts. An example of such a task could be the recognition of cancer cells in medical images. In order to support physicians in this task with an AI, training examples must first be provided by trained personnel. These training examples are images in which cancer cells have been identified and marked by the experts. The AI can then be trained with these examples, i.e., the AI learns to classify images automatically. The goal of the dissertation is to limit the selection of training examples to the most important ones and thus save costs. For this purpose, Dr. Daniel Kottke has developed a mathematical approach that maximizes the goodness of the AI while taking its uncertainties into account.

Daniel Kottke received his PhD in 2021 from the Department of Electrical Engineering/Computer Science at the University of Kassel (Department of Intelligent Embedded Systems, headed by Prof. Dr. Bernhard Sick). At this department he currently leads the team "Methods for Intelligent Interactive Systems".

Each year, the Classification Society presents a Distinguished Dissertation Award for a dissertation on clustering, classification, or related areas of data analysis, including associated theory and/or applications. The Classification Society's Distinguished Dissertation Award is sponsored by Springer, an academic publisher.

Contact:
Dr. Daniel Kottke
Team Leader: Methods for Intelligent Interactive Systems (I2S)
Phone: +49 561 804-6036
Email: daniel.kottke[at]uni-kassel[dot]de