The section for Cognitive System at DTU Compute is hosting a Kaggle Workshop from the 12th to the 14th of March 2015. We have invited 3 of the best data-scientists in the world: Maxim Milakov, Sander Dieleman and Abhishek Thakur, all Kaggle Masters.
Maxim Milakov is a researcher in machine learning and high-performance computing. Maxim is currently employed at NVIDIA. He has designed and implemented nnForge – a library for training convolutional and fully connected neural networks, with CPU and GPU (CUDA) backends.
Sander Dieleman is a PhD student in the Reservoir Lab at Ghent University in Belgium. His research is on learning hierarchical representations of musical audio signals for classification and recommendation, with a focus on deep learning and feature learning.
Abhishek Thakur is a PhD student at Universität Paderborn, Germany where he is working in the Analytic Information Systems and Business Intelligence group. Abhishek has participated in not less than 74 Kaggle competitions giving him a current rank as the 9th best Kaggler in the world!
Angel Diego Cuñado Alonso is a “Highly non-linear Machine Learning Engineer” at Tradeshift. He has worked in multidisciplinary research and development teams connected to the Business, IT and Bio-Tech fields. In 2014 he was part of hosting the Tradeshift Text Classification challenge on Kaggle.
On the first day of the workshop (Thursday the 12th of March), the four invited speakers will be giving talks about their specific fields of expertise within data science, and about how they continue to outperform others in data science competitions on Kaggle.
On the second day of the workshop (Friday the 13th of March), all participants in the workshop will work together on solving a Kaggle challenge (we will have selected one before the workshop). The Kaggle masters will be guiding our effort, and showing how to quickly and efficiently attack a practical problem in data science.
For those who are interested, some of us will continue the work on the challenge on Saturday the 14th of March.
Participation: If you are not from the Cognitive Systems section at DTU Compute and you wish to participate in the workshop, please notify in an e-mail to firstname.lastname@example.org first.
Thursday the 12th of March
Location: Technical University of Denmark. Building 306, Auditorim 031.
10:00-10:20: Introduction to Kaggle and the workshop by David Kofoed Wind
10:20-11:15: What have I learned by competing on Kaggle by Abhishek Thakur
11:15-12:00: Deep Learning with GPU’s by Maxim Milakov
13:00-13:45: Applying ConvNets to Kaggle Competitions by Maxim Milakov
13:45-14:45: Galaxy Zoo and Whale Detection Challenge by Sander Dieleman
14:45-15:10: Deep Learning for Music by Sander Dieleman
15:10-15:40: Lessons Learned by Hosting a Competition by Angel Diego Cuñado Alonso
Friday the 13th of March
Location: Technical University of Denmark. Building 321, first floor in the lab.
10:00-12:00 Work on challenge
13:00-??:?? Work on challenge
Saturday the 14th of March
Location: To be announced
10:00-??:?? Work on challenge