דילוג לניווט ראשי דילוג לחיפוש דילוג לתוכן הראשי

Greed Is Good: Near-Optimal Submodular Maximization via Greedy Optimization

  • Moran Feldman
  • , Christopher Harshaw
  • , Amin Karbasi

פרסום מחקרי: פרסום בכתב עתמאמר מכנסביקורת עמיתים

תקציר

It is known that greedy methods perform well for maximizing monotone submodular functions. At the same time, such methods perform poorly in the face of non-monotonicity. In this paper, we show—arguably, surprisingly—that invoking the classical greedy algorithm O(k)-times leads to the (currently) fastest deterministic algorithm, called REPEATEDGREEDY, for maximizing a general submodular function subject to k-independent system constraints. REPEATEDGREEDY achieves (1 + O(1/k))k approximation using O(nrk) function evaluations (here, n and r denote the size of the ground set and the maximum size of a feasible solution, respectively). We then show that by a careful sampling procedure, we can run the greedy algorithm only once and obtain the (currently) fastest randomized algorithm, called SAMPLEGREEDY, for maximizing a submodular function subject to k-extendible system constraints (a subclass of k-independent system constrains). SAMPLEGREEDY achieves (k + 3)-approximation with only O(nr/k) function evaluations. Finally, we derive an almost matching lower bound, and show that no polynomial time algorithm can have an approximation ratio smaller than k +1/2 - ε. To further support our theoretical results, we compare the performance of REPEATEDGREEDY and SAMPLEGREEDY with prior art in a concrete application (movie recommendation). We consistently observe that while SAMPLEGREEDY achieves practically the same utility as the best baseline, it performs at least two orders of magnitude faster.

שפה מקוריתאנגלית
עמודים (מ-עד)758-784
מספר עמודים27
כתב עתProceedings of Machine Learning Research
כרך65
סטטוס פרסוםפורסם - 2017
אירוע30th Conference on Learning Theory, COLT 2017 - Amsterdam, הולנד
משך הזמן: 7 יולי 201710 יולי 2017

הערה ביבליוגרפית

Publisher Copyright:
© 2017 M. Feldman, C. Harshaw & A. Karbasi.

טביעת אצבע

להלן מוצגים תחומי המחקר של הפרסום 'Greed Is Good: Near-Optimal Submodular Maximization via Greedy Optimization'. יחד הם יוצרים טביעת אצבע ייחודית.

פורמט ציטוט ביבליוגרפי