The Mathematics of Machine Learning Group, led by Philipp Petersen at the University of Vienna, is currently inviting applications for a three-year PhD student position in the framework of the FWF project: ”Structured Singularities in Deep Learning”. In deep learning, structured singularities arise, for example, as decision boundaries in a classification problem. Problems where structured singularities are present form some of the few problem cases where deep neural networks are provably superior to all classical machine learning techniques. It can be shown that certain parameters, such as the regularity or size of the decision boundary, as well as related problems, such as the margin and the interplay of noise with the boundary, have a quantifiable effect on the learnability of a problem, as well as on how well-learned classifiers generalize. In this project, these singularities and their effect on the learnability of functions will be studied in a principled way. The position has a work-load of 30hrs/week and has no teaching duties. The salary is in accordance with collective labor agreements; see https://personalwesen.univie.ac.at/en/jobs-recruiting/job-center/salary-scheme/. The prefered starting date is in Autumn/Winter 2023. We are seeking a mathematician with a Master’s degree or equivalent, who is enthusiastic about collaborating with project partners at the University of Tokyo, including regular meetings and potential research visits. Candidates with a background in machine learning, probability theory, functional analysis, or applied harmonic analysis will be given preference. Furthermore, sound coding skills, especially in Python, are desirable. Interested applicants are invited to submit their application, preferably in a single PDF file. The application should include a cover letter, CV, and contact details for two references. Applications should be sent to philipp.petersen@univie.ac.at. Although the position will remain open until filled, to ensure thorough consideration, we recommend submitting your application before July 15th, 2023. The University pursues a non-discriminatory employment policy and values equal opportunities as well as diversity. The University puts special emphasis on increasing the number of women in academic positions. Given equal qualification, preference will be given to female applicants. |