| layout | title | permalink |
|---|---|---|
page |
About |
/about/ |
Hey there, I'm Theodoros (Theodore) Vasiloudis and I'm a Machine Learning researcher at the Swedish Institute for Computer Science in Stockholm. My main research interests include large-scale machine learning, online learning, and decision trees. I also contribute to the Apache Flink distributed processing engine, and am one of the core developers of its Machine Learning library, FlinkML.
I completed my Master's Degree in Machine Learning at the Royal Institute of Technology, KTH. My thesis work was performed at Spotify AB, under the supervision of Hedvig Kjellström, KTH and Boxun Zhang, Spotify. My work centered on context-aware recommendation systems, and investigating how do contextual variables such as time affect user choice. You can read my thesis here.
I am a graduate of Aristotle University of Thessaloniki, school of Computer Science. During my studies I followed the Software Engineering track and focused on Machine Learning and Data Science courses which lead to a thesis on the subject of multi-label classification using Bayes networks, working with the Machine Learning and Knowledge Discovery group of the school, under the supervision of Dr. Grigorios Tsoumakas.
I'm industrial Phd student at KTH, and I try to complete one internship every year of my PhD. My work on Flink was done while in Berlin with Data Artisans in the summer of 2015, the following year I worked on streaming session length prediction at Pandora Media, while in 2017 I interned at Amazon, working on GANs for NLP for Alexa.