Előadó: Juhász Péter (University of Oxford)
Az előadás témája: Machine learning applications for ultracold atoms & the Oxford Bose–Einstein condensate experiment
Az előadás időpontja: 2021. január 26., kedd, 10.00
Helyszín: videokonferencia, https://letsmeet.wigner.hu/szeminarium
Bose–Einstein condensates (BECs) have been a very successful vehicle to study quantum phenomena in the past twenty years. In Oxford, we are currently in the process of building an erbium BEC machine, which will be used to study the effect of its high magnetic dipole moment on quantum dynamics. In this talk, I will describe the experiment in detail (the trapping and cooling techniques we use – a Zeeman slower, a narrow-line magneto-optical trap, an optical dipole trap and an optical box trap) and some of the research themes we would like to explore (supersolidity and the 'roton spectrum', non-Markovian reservoirs for quantum computers and quantum phase transition dynamics). I will finally touch on the machine learning techniques that we currently use to help data processing and that we plan to introduce soon to optimise the experiment.