{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/sampling-from-a-k-dpp-without-looking-at-all-1","title":"Sampling from a k-DPP without looking at all items","arxiv_id":null,"date":"2020-12-01","proceeding":"NeurIPS 2020 12","authors":["Daniele Calandriello","Michal Derezinski","Michal Valko"],"abstract":"Determinantal point processes (DPPs) are a useful probabilistic model for selecting\na small diverse subset out of a large collection of items, with applications in\nsummarization, recommendation, stochastic optimization, experimental design and\nmore. Given a kernel function and a subset size k, our goal is to sample k out\nof n items with probability proportional to the determinant of the kernel matrix\ninduced by the subset (a.k.a. k-DPP). Existing k-DPP sampling algorithms require\nan expensive preprocessing step which involves multiple passes over all n items,\nmaking it infeasible for large datasets. A naïve heuristic addressing this problem is\nto uniformly subsample a fraction of the data and perform k-DPP sampling only on\nthose items, however this method offers no guarantee that the produced sample will\neven approximately resemble the target distribution over the original dataset. In this\npaper, we develop alpha-DPP, an algorithm which adaptively builds a sufficiently large uniform\nsample of data that is then used to efficiently generate a smaller set of k items,\nwhile ensuring that this set is drawn exactly from the target distribution defined\non all n items. We show empirically that our\nalgorithm produces a k-DPP sample after observing only a small fraction of all\nelements, leading to several orders of magnitude faster performance compared to\nthe state-of-the-art. Our implementation of alpha-DPP is provided at https://github.com/guilgautier/DPPy/.","url_abs":"http://proceedings.neurips.cc/paper/2020/hash/4d410063822cd9be28f86701c0bc3a31-Abstract.html","url_pdf":"http://proceedings.neurips.cc/paper/2020/file/4d410063822cd9be28f86701c0bc3a31-Paper.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"sampling-from-a-k-dpp-without-looking-at-all-1","repo_url":"https://github.com/guilgautier/DPPy","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"all","task_name":"All"},{"task_slug":"experimental-design","task_name":"Experimental Design"},{"task_slug":"point-processes","task_name":"Point Processes"},{"task_slug":"stochastic-optimization","task_name":"Stochastic Optimization"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}