I am a senior research scientist at RISE AI (Research institutes of Sweden) and the leader of our machine learning research group in Gothenburg.
I have a PhD in machine learning from Chalmers university of technology, my thesis is called Representation learning for natural language (click for details). I have worked on summarization, dialogue systems, adversarial training, and character-based RNNs for sequence labeling and morphological analogies. I also have experience in convnets and policy gradient methods.
I enjoy doing AI research to make the world a better place for us all, and I am an avid listener of the Talking Machines podcast every two weeks.
I tweet and blog about things I find interesting, and give talks in our popular learning machines seminar series at RISE AI (and at other events). During 2016-2017, I was the organizer of a seminar series at the machine learning research group at Chalmers. Most of my recent code is written in Python with Pytorch, and I also have experience with Tensorflow, Theano, and Lasagne.
Also see my licentiate thesis, titled "Multi-document summarization and semantic relatedness", and my master's thesis Dynamics of geographical routing in small-world networks.
I have developed and taught a PhD course in deep learning, together with Mikael Kågebäck and Fredrik Johansson. The course covered some of the most important topics in the field, including convolutional neural networks, recurrent neural networks, unsupervised methods, and regularization techniques. We used a “flipped classroom” approach, with video lectures from some of the best researchers in the field, along with discussion sessions.
I have supervised a number of master students.
Furthermore, I have taught the following courses.
During 2016-217, I was the organizer of Chalmer Machine Learning Seminars.
When not doing research or teaching, I am a long distance runner and the lucky father of two wonderful children.