|COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring.|
Using Context and Insight for the Analysis of LittleData?
If you have a question about this talk, please contact Zoubin Ghahramani.
We live in the era of Big Data with many interesting challenges for Machine Learning. In the opposite regime of “Little Data”, a limited population size, limited measurement time or a limited budget constrains the amount of data available. Examples include magnetic resonance imaging, microarray experiments or portfolio management. To obtain sensible predictions it is then crucial to take modeling assumptions into account. In this talk I will cover challenges, both computational and conceptual, that come up in the presence of a limited amount of data and how they can be overcome. I will focus on causality and high-dimensional covariance estimation and present a new algorithm that allows to compute an asymptotically unbiased, sparse and positive semidefinite estimator for covariance matrices using the SCAD penalty from high-dimensional statistics.
This talk is part of the Machine Learning @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
Other listsCambridge Carbon Nanotechnology Society Department of Psychiatry talks stream Joint Machine Learning Seminars
Other talksOverview of cancer theoretical framework Universal properties of quantum gravity Ash, Ice, Mud: Tephras and Rapid Climate Changes Call My Bluff Trapping Mink for Water Vole recovery in the upper Cam catchment How quantum physics democratized music