Machine learning seminar – Separating Higgses and interpreting probability, and cellular stress after lack of sleep
March 27 @ 14:15 - 16:00
Welcome to the joint HVL, UiB and HUS machine learning seminar!
The seminar series is an informal platform where people interested in machine learning can meet and discuss across institutions and disciplines. Experienced users, newcomers, and everyone in between are welcome to participate. In this way, we can share experiences, learn from each other and maybe develop a foundation for future collaborations.
Time: Wednesday March 27th 2019,14:15-16:00
Place: Auditorium 11 (E106), HVL – Campus Kronstad (directions here)
Steffen Mæland, UiB: Deep learning: Separating Higgses and interpreting probability.
Abstract: Machine learning is widely used within particle physics, most commonly for signal vs background classification. In this talk, a challenging signal vs signal scenario is discussed, specifically a two-Higgs model with mass-degenerate Higgses, where classification is complicated by nearly indistinguishable observables. The behaviour of a deep neural network is discussed, and lastly, some food for thought about the probabilistic interpretation of any machine learning algorithm is given. The talk will not be very technical, focusing mostly on intuition.
Alvhild Alette Bjørkum, HVL: Cellular stress after lack of sleep – how to measure it including the use of machine learning
Abstract: Sleep problems in daily life – in health and disease – is often so that we get too little of it or it is too skewed and thereby of bad quantity and quality and mistimed. We have some understanding about what lack of sleep and skewed sleep do to sickness and health, learning and mental capacity and stability. However why we sleep is still an unrevealed mystery. How we measure sleep and the affects of lack of it depends on our methods, our interpretation of our findings either it is quantitative or qualitative data and our methods must be validated. We also need to keep track of our circadian rhythm and together with sleep it is of uttermost importance for many bodily functions that these rhythms are synchronized. In shiftwork, with travel across time zones and in our social life and demanding 24/7- around the clock activities – we falter ad synchronization. In 2017 the Nobel prize in physiology and medicine was given to the researchers who discovered the cellular machinery of the circadian clock. Sleep-research have by this been actualized, since both sleep/wake-rhythms and circadian rhythms are doomed to tangle. Some of us (Bjørkum) in Bergen have for years used learning algorithms, neuronal network e.g. on our measurement of cellular changes as at the protein level in bodily fluids as blood serum and saliva/spit to try to understand compromised cellular activities and hunt for biomarkers therein. Also, our international collaborators we have had for years and have bilateral agreements with at Harvard (I worked there for 4 yrs in two different groups) and Pennsylvania University have started using machine learning on their sleep-data, as presented at the World Sleep meeting in Prague in 2018. We recently have teamed up in a larger sleep and chronobiology-network here in Bergen. Read more about us in BeSCN here: https://www.uib.no/en/rg/sc. Also check out our work on tissue engineering and cellular stress the latest 5-6 years: https://uhnettvest.no/finner-nye-metoder-for-a-kartlegge-biologiske-effekter-av-nanopartikler/
If you are interested in giving a talk at some of the upcoming meetings, please email us (Therese.Berge.Sjursen@hvl.no).