[Ml-stat-talks] Wilks Seminar this Friday
Amir Ali Ahmadi
a_a_a at princeton.edu
Tue Mar 10 17:48:42 EDT 2015
The talk in ORFE's stats seminar this Friday may be of interest to many of you.
Date and time: Friday, March 13, 12:30 PM
Location: Sherrerd 101
A Tale of Two Norms
* Nathan Srebro
* Toyota Technological Institute at Chicago
There has been much interest in recent years in various ways of constraining the complexity of matrices based on factorizations into a product of two simpler matrices. Such measures of matrix complexity can then be used as regularizers for such tasks as matrix completion, collaborative filtering, multi-task learning and multi-class learning. In this talk I will discuss two forms of matrix regularization which constrain the norm of the factorization, namely the trace-norm (aka nuclear-norm) and the so-called max-norm (aka γ2:ℓ1→ℓ∞γ2:ℓ1→ℓ∞γ2:ℓ1→ℓ∞ norm). I will both argue that they are independently motivated and often better model data then rank constraints, as well as explore their relationships to the rank. In particular, I will discuss how simple low-rank matrix completion guarantees can be obtained using these measures, and without various "incoherence" assumptions. I will present both theoretical and empirical arguments for why the max-norm might actually be a better regularizer, as well as a better convex surrogate for the rank.
Based on joint work with Rina Foygel, Jason Lee, Ben Recht, Russ Salakhutdinov, Ohad Shamir, Adi Shraibman and Joel Tropp and others.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ml-stat-talks