AIC Seminar Series
On Learning from Data using (un-)informative Priors
Notice: Hosted by David Israel. WebEx at sri-meetings.webex.com, number 629 890 179, sound at 1-888-355-1249, number 749045
Date: 2013-06-27 at 16:00
Location: EJ228 (SRI E building) (Directions)
In Bayesian statistics, the maximum-a-posteriori estimate is obtained
from the prior combined with the likelihood. In this talk, I will
discuss, in two examples, how one can choose an (un-)informative
Bayesian prior as to influence the obtained result in (un-)expected
The first part of this talk is concerned with the Bayesian approach to
structure learning of Bayesian networks. The equivalent sample size
(ESS) in the Dirichlet prior over the model parameters is often
interpreted as the strength of prior belief, relative to the weight of
the training data. In this talk, I will show how the chosen ESS value
affects the number of edges in the maximum-a-posteriori estimate of
the Bayesian network structure in an unexpected way–from the empty
graph to the complete graph.
The second part of this talk is concerned with learning recommender
systems that can make personalized suggestions, e.g., of movies, based
on a user’s past viewing history. The objective of our recommender
system is to rank movies according to each user’s preferences. Users’
preferences can be estimated from their feedback data like plays or
ratings of movies, among others. While the extreme data sparsity in
the collected data is a well-known problem (each user interacts with
only a small fraction of all movies), I will focus on the problem that
the feedback data are missing not at random (MNAR). The MNAR nature
of the feedback data originates from the fact that users can choose
which movies to play or to rate. I will present a simple yet
effective approach to tackle this problem using an informative prior
with a few parameters that can be determined empirically.
Harald Steck is a data scientist at Netflix, where he develops
personalization algorithms and recommender systems. He has over ten
years of experience in machine learning, in particular in graphical
models and more recently in recommender systems. He has conducted
research at various industrial and academic organizations, including
Bell Labs, ETH Zurich in Switzerland, MIT AI Lab, as well as Technical
University of Munich in Germany, where he obtained a PhD degree in
Computer Science in 2001.
Please arrive at least 10 minutes early as you will need to sign in by
following instructions by the lobby phone at Building E. SRI is located
at 333 Ravenswood Avenue in Menlo Park. Visitors may park in the parking
lots off Fourth Street. Detailed directions to SRI, as well as maps, are
available from the Visiting AIC web page.
There are two entrances to SRI International located on Ravenswood Ave.
Please check the Builing E entrance signage.
©2014 SRI International 333 Ravenswood Avenue, Menlo Park, CA 94025-3493