Timezone: »

A Non-generative Framework and Convex Relaxations for Unsupervised Learning
Elad Hazan · Tengyu Ma

Mon Dec 05 09:00 AM -- 12:30 PM (PST) @ Area 5+6+7+8 #45

We give a novel formal theoretical framework for unsupervised learning with two distinctive characteristics. First, it does not assume any generative model and based on a worst-case performance metric. Second, it is comparative, namely performance is measured with respect to a given hypothesis class. This allows to avoid known computational hardness results and improper algorithms based on convex relaxations. We show how several families of unsupervised learning models, which were previously only analyzed under probabilistic assumptions and are otherwise provably intractable, can be efficiently learned in our framework by convex optimization.

Author Information

Elad Hazan (Princeton University and Google Brain)
Tengyu Ma (Princeton University)

More from the Same Authors