Supervised and unsupervised learning are two fundamental learning schemes whose difference lies in the presence and absence of a supervisor (i.e. entity which provides examples) respectively. On the other hand, transfer learning aims at improving the learning of a task by using auxiliary knowledge. The goal of this thesis was to investigate how the two fundamental paradigms, supervised and unsupervised learning, can collaborate in the setting of transfer learning. As a result, we developed transfer k-means, a transfer learning variant of the popular k-means algorithm.
The proposed method enhances the unsupervised nature of k-means, using supervision from a different but related context as a seeding technique, in order to improve the algorithm’s performance towards more meaningful results. We provide approximation guarantees based on the nature of the input and we experimentally validate the benefits of the proposed method using documents as a real-world example.
Pelagia Teloni has graduated from the Department of Informatics and Telecommunications of University of Athens and holds a master’s degree in Logic, Algorithms and Computation from the MPLA (http://mpla.math.uoa.gr/en/) graduate Program sponsored by the Departments of Mathematics, Informatics and M.I.TH.E. (Methodology, History and Theory of Science), of the University of Athens, by the General Sciences Department and the Department of Electrical and Computer Engineering of the National Technical University of Athens and by the Department of Computer Engineering and Information of the University of Patras. She is currently working as a Machine Learning Engineer at Eyequant (www.eyequant.com) where she is mainly working on Saliency Prediction powered by Machine/Deep Learning using eye-tracking data.