Timezone: »

 
Poster
Minimizing a Submodular Function from Samples
Eric Balkanski · Yaron Singer

Wed Dec 06 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #156
in Posters Wed »

In this paper we consider the problem of minimizing a submodular function from training data. Submodular functions can be efficiently minimized and are conse- quently heavily applied in machine learning. There are many cases, however, in which we do not know the function we aim to optimize, but rather have access to training data that is used to learn the function. In this paper we consider the question of whether submodular functions can be minimized in such cases. We show that even learnable submodular functions cannot be minimized within any non-trivial approximation when given access to polynomially-many samples. Specifically, we show that there is a class of submodular functions with range in [0, 1] such that, despite being PAC-learnable and minimizable in polynomial-time, no algorithm can obtain an approximation strictly better than 1/2 − o(1) using polynomially-many samples drawn from any distribution. Furthermore, we show that this bound is tight using a trivial algorithm that obtains an approximation of 1/2.

Keywords: Submodular Optimization · Hardness of Learning and Approximations
[ Paper] 

Author Information

Eric Balkanski (Harvard University)
Yaron Singer (Harvard University)

More from the Same Authors