BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Kernel density estimation techniques and their applications in ML
DTSTART:20240705T140000
DTEND:20240705T160000
DTSTAMP:20260525T190836Z
UID:23327267e06fc5a6c0caa75ab770867254cf296d15dd4c3ddf484629
CATEGORIES:Conferences - Seminars
DESCRIPTION:Ekaterina Kochetkova\nEDIC candidacy exam\nExam president: Pro
 f. Mika Göös\nThesis advisor: Prof. Michael Kapralov\nCo-examiner: Prof.
  Ola Svensson\n\nAbstract\nThe Kernel Density Estimation (KDE) problem fin
 ds applications in many practical tasks involving data analysis and approx
 imate fast computations. Naturally\, it has become important to machine le
 arning\, and attracted great research interest. We present two papers show
 ing how different techniques from this line of work can be applied to obta
 in practical tools for training large language models\, as well as one rec
 ent paper providing a result on a problem used as a building block in a pa
 rticular approach to the KDE problem.\n\nBackground papers\n\n	Janardhan K
 ulkarni\, Victor Reis\, Thomas Rothvoss. "Optimal Online Discrepancy Minim
 ization." https://arxiv.org/pdf/2308.01406\n	Moses Charikar\, Beidi Chen\
 , Christopher Re\, Erik Waingarten. "Fast Algorithms for a New Relaxation 
 of Optimal Transport". Proceedings of Machine Learning Research vol 195:1
 –32\, 2023. https://proceedings.mlr.press/v195/charikar23a/charikar23a.
 pdf\n	Josh Alman\, Zhao Song. "Fast Attention Requires Bounded Entries". 
 37th Conference on Neural Information Processing Systems (NeurIPS 2023). 
 https://proceedings.neurips.cc/paper_files/paper/2023/file/c72861451d6fa9d
 fa64831102b9bb71a-Paper-Conference.pdf\n
LOCATION:BC 329 https://plan.epfl.ch/?room==BC%20329
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR