Seminar/workshop, Lecture • Białobrzegi
#NGSchool2019 Autumn School in Machine Learning for Biomedicine
#NGSchool2019 will be an 8-day intensive autumn school. To implement best strategy into transferring knowledge of something as complex as ML implementation in biomedicine we dived the school into two modules - taught and practical. First 2 days will be entirely dedicated to introductory sessions in programming languages: R and Python, unix command line environment and bash scripting as well as probability theory and statistics. As we are selecting participants from diverse backgrounds, introductory sessions will bring everybody to the level necessary to fully benefit from the course. The participants will also get familiarized with high-computing cluster (HPC) environments, which they will use during the hackathons and continue using later for efficient analysis of data generated in high-throughput sequencing experiments. The following 3 days will be entirely dedicated to hands-on workshops and lectures on various areas of statistical and machine learning methods such as regression, classification, models, boosting, neural networks, unsupervised learning, and other related topics. The following days will be reserved for hackathons. Workshops are scheduled for 3 hours each, while hackathons will last two days. We plan to organize 5 parallel hackathons - with 6 participants per group. Each hackathon, guided by the mentor, will tackle a real-world problem by applying the knowledge gained throughout the taught component of the school to establish a working executable protocol. This will result in ready-to-use solutions for very focused sets of problems. We will publish the code and tutorials via our website, github repository and academic publication.
As we strive to improve, we learned from the last edition that good quality accommodation, as well as relying on fewer but more versatile subcontractors is essential for an efficient and productive workshop. In 2018, we relied on university-owned dorms and multiple subcontractors to reduce costs. However, this proved complicated, put a lot of burden on the organising team and resulted in a few complains from the participants. Therefore, starting from 2019, we’ll take into account not only how competitive in terms of price the potential subcontractor’s offer is, but also their excellence and comprehensiveness. That’s why for the venue this year we chose not the cheapest offer, but in our judgement, the best venue that (1) provided all the necessary components that would allow to organize a professional event, (2) was close to communication nodes and (3) was vouched for by one of our partners as extremely reliable.
Thanks to its location and easy access from Warsaw, a collaboration with the aforementioned venue will also allow us to organize an event open to the scientific community. We plan to host a panel on the ethics in the Machine Learning and Artificial Intelligence. This event, engaging the participants and the scientific community from Warsaw's research institutions will provide an unparalleled opportunity for an educated debate and, followed by a networking event, will hopefully allow for a more involved and informal dialog that will continue long after the summer school. We plan to engage media specialised in science popularisation, allowing the debate to reach an even broader audience through publications. On top of that, we will be streaming the debate online, allowing for participation of the online community.
https://ngschool.eu/2019
