Composable Core-sets for Determinant Maximization: A Simple Near-Optimal Algorithm

[edit]

Sepideh Mahabadi, Piotr Indyk, Shayan Oveis Gharan, Alireza Rezaei ;
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:4254-4263, 2019.

Abstract

“Composable core-sets” are an efficient framework for solving optimization problems in massive data models. In this work, we consider efficient construction of composable core-sets for the determinant maximization problem. This can also be cast as the MAP inference task for “determinantal point processes", that have recently gained a lot of interest for modeling diversity and fairness. The problem was recently studied in \cite{indyk2018composable}, where they designed composable core-sets with the optimal approximation bound of $O(k)^k$. On the other hand, the more practical “Greedy" algorithm has been previously used in similar contexts. In this work, first we provide a theoretical approximation guarantee of $C^{k^2}$ for the Greedy algorithm in the context of composable core-sets; Further, we propose to use a “Local Search" based algorithm that while being still practical, achieves a nearly optimal approximation bound of $O(k)^{2k}$; Finally, we implement all three algorithms and show the effectiveness of our proposed algorithm on standard data sets.

Related Material