Simon Kamronn, Andreas Trier Poulsen, and Lars Kai Hansen. 2015. Multiview Bayesian correlated component analysis. Neural Computation 2015 27:10, 2207-2230.
Focus has during the masters, and especially in the PhD, been on dealing with and modelling large, noisy data. I have been gathering data pervasively from study participants using smartphones, predicting activity and domain with neural networks, and modelling causal intervention effects with Bayesian models in probabilistic programming languages such as Stan which enables flexibility and exact inference through sampling, and Edward which enables inference with millions of samples using stochastic variational inference.
- Ph.D in Machine Learning/Data Science, Technical University of Denmark, Cognitive Systems, 2018
- M.Sc. in Medicine and Technology, Technical University of Denmark, 2014
- B.Sc. in Medicine and Technology, Technical University of Denmark, 2011
- Co-founder, Siren IVS, www.siren.care, Copenhagen, 2015.
- Developing a temperature sensings sock for prevention of diabetic foot ulcers. Worked for 6 months on the project but left to pursue a PhD.
- Research Assistant, DTU Compute, Kgs. Lyngby, 2014 - 2015.
- Working on the project Neuro 24/7, long term ear-EEG monitoring. Smartphone sensor capture and processing, speaker recognition, EEG experiments
- Co-founder, Cortrium, www.cortrium.com, Copenhagen, 2013 - 2015.
- Working on ICA de-noising, feature extraction and classification of two-lead wireless EEG
- Software Developer Consultant in LabVIEW, Medimatic A/S, Hellerup, 2011 - 2017.
- Developed a program for data acquisition, organisation and processing to work with a distal blood pressure measurement device in hospitals.
- Demonstration, sales and technical support in most of the major hospitals in Denmark.
Andreas Trier Poulsen, Simon Kamronn, Jacek Dmochowski, Lucas C Parra, Lars Kai Hansen. 2017. EEG in the classroom: Synchronised neural recordings during video presentation. Scientific Reports 7.
Marco Fraccaro, Simon Kamronn, Ulrich Paquet, Ole Winther. 2017. A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised Learning. Advances in Neural Information Processing Systems 30.
Talk at NIPS, Los Angeles, California
- Numpy, Pandas, BColz, Dask, Bokeh
- Stan, Edward, PyMC3
- Machine Learning
- TensorFlow, PyTorch, Scipy, Scikit-learn