Material Detail

A Framework for Probability Density Estimation

A Framework for Probability Density Estimation

This video was recorded at NIPS Workshop on Representations and Inference on Probability Distributions, Whistler 2007. The talk introduces a new framework for learning probability density functions by assessing their performance against a set of 'test functions'. A theoretical analysis suggests that we can tailor a distribution for a class of tasks by training it to fit a small subsample. There is a trade-off between the complexity of the class of distributions used for the distribution and the complexity of the set of tasks. Experimental evidence is given to support the theoretical analysis.


  • User Rating
  • Comments
  • Learning Exercises
  • Bookmark Collections
  • Course ePortfolios
  • Accessibility Info

More about this material


Log in to participate in the discussions or sign up if you are not already a MERLOT member.