In this paper we consider the following maximum budgeted allocation<\/i> (MBA) problem: Given a set of $m$ indivisible items and $n$ agents, with each agent $i$ willing to pay $b_{ij}$ on item $j$ and with a maximum budget of $B_i$, the goal is to allocate items to agents to maximize revenue. The problem naturally arises as auctioneer revenue maximization in budget-constrained auctions and as the winner determination problem in combinatorial auctions when utilities of agents are budgeted-additive. Our main results are as follows: (i) We give a $3\/4$-approximation algorithm for MBA improving upon the previous best of $\\simeq0.632$ [N. Andelman and Y. Mansour, Proceedings of the <\/i>9th Scandinavian Workshop on Algorithm Theory (SWAT)<\/i>, 2004, pp. 26-38], [J. Vondr\u00e1k, Proceedings of the <\/i>40th Annual ACM Symposium on the Theory of Computing (STOC)<\/i>, 2008, pp. 67-74] (also implied by the result of [U. Feige and J. Vondr\u00e1k, Proceedings of the <\/i>47th IEEE Symposium on Foundations of Computer Science (FOCS)<\/i>, 2006, pp. 667-676]). Our techniques are based on a natural LP relaxation of MBA, and our factor is optimal in the sense that it matches the integrality gap of the LP. (ii) We prove it is NP-hard to approximate MBA to any factor better than $15\/16$; previously only NP-hardness was known [T. Sandholm and S. Suri, Games Econom. Behav.<\/i>, 55 (2006), pp. 321-330], [B. Lehmann, D. Lehmann, and N. Nisan, Proceedings of the <\/i>3rd ACM Conference on Electronic Commerce (EC)<\/i>, 2001, pp. 18-28]. Our result also implies NP-hardness of approximating maximum submodular welfare with demand oracle<\/i> to a factor better than $15\/16$, improving upon the best known hardness of $275\/276$ [U. Feige and J. Vondr\u00e1k, Proceedings of the <\/i>47th IEEE Symposium on Foundations of Computer Science (FOCS)<\/i>, 2006, pp. 667-676]. (iii) Our hardness techniques can be modified to prove that it is NP-hard to approximate the generalized assignment problem<\/i> (GAP) to any factor better than $10\/11$. This improves upon the $422\/423$ hardness of [C. Chekuri and S. Khanna, Proceedings of the <\/i>11th Annual ACM-SIAM Symposium on Discrete Algorithms (SODA)<\/i>, 2000, pp. 213-222], [M. Chleb\u00edk and J. Chleb\u00edkov\u00e1, Proceedings of the <\/i>8th Scandinavian Workshop on Algorithm Theory (SWAT)<\/i>, 2002, pp. 170-179]. We use iterative rounding<\/i> on a natural LP relaxation of the MBA problem to obtain the $3\/4$-approximation. We also give a $(3\/4-\\epsilon)$-factor algorithm based on the primal-dual schema which runs in $\\tilde{O}(nm)$ time, for any constant $\\epsilon>0$.<\/p>\n","protected":false},"excerpt":{"rendered":"
In this paper we consider the following maximum budgeted allocation (MBA) problem: Given a set of $m$ indivisible items and $n$ agents, with each agent $i$ willing to pay $b_{ij}$ on item $j$ and with a maximum budget of $B_i$, the goal is to allocate items to agents to maximize revenue. The problem naturally arises […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":null,"msr_publishername":"Society for Industrial and Applied Mathematics Philadelphia, PA, USA","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"SIAM Journal on Computing","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"6","msr_journal":"SIAM Journal on Computing","msr_number":"","msr_organization":"","msr_pages_string":"2189-2211","msr_page_range_start":"2189","msr_page_range_end":"2211","msr_series":"","msr_volume":"39","msr_copyright":"","msr_conference_name":"","msr_doi":"10.1137\/080735503","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":"","msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2010-03-01","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":0,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13561],"msr-publication-type":[193715],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-324356","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-algorithms","msr-locale-en_us"],"msr_publishername":"Society for Industrial and Applied Mathematics Philadelphia, PA, USA","msr_edition":"SIAM Journal on Computing","msr_affiliation":"","msr_published_date":"2010-03-01","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"2189-2211","msr_chapter":"","msr_isbn":"","msr_journal":"SIAM Journal on Computing","msr_volume":"39","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"6","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"324359","msr_publicationurl":"","msr_doi":"10.1137\/080735503","msr_publication_uploader":[{"type":"file","title":"On the Approximability of Budgeted Allocations and Improved Lower Bounds for Submodular Welfare Maximization and GAP","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2016\/11\/On-the-Approximability-of-Budgeted-Allocations-and-Improved-Lower-Bounds-for-Submodular-Welfare-Maximization-and-GAP.pdf","id":324359,"label_id":0},{"type":"doi","title":"10.1137\/080735503","viewUrl":false,"id":false,"label_id":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[],"msr-author-ordering":[{"type":"user_nicename","value":"dechakr","user_id":31593,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=dechakr"},{"type":"text","value":"Gagan Goel","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"article","related_content":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/324356","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/324356\/revisions"}],"predecessor-version":[{"id":525606,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/324356\/revisions\/525606"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=324356"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=324356"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=324356"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=324356"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=324356"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=324356"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=324356"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=324356"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=324356"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=324356"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=324356"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=324356"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=324356"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}