{"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/mission-ultra-large-scale-feature-selection","title":"MISSION: Ultra Large-Scale Feature Selection using Count-Sketches","arxiv_id":"1806.04310","date":"2018-06-12","proceeding":null,"authors":["Amirali Aghazadeh","Ryan Spring","Daniel Lejeune","Gautam Dasarathy","Anshumali Shrivastava","Richard G. Baraniuk"],"abstract":"Feature selection is an important challenge in machine learning. It plays a\ncrucial role in the explainability of machine-driven decisions that are rapidly\npermeating throughout modern society. Unfortunately, the explosion in the size\nand dimensionality of real-world datasets poses a severe challenge to standard\nfeature selection algorithms. Today, it is not uncommon for datasets to have\nbillions of dimensions. At such scale, even storing the feature vector is\nimpossible, causing most existing feature selection methods to fail.\nWorkarounds like feature hashing, a standard approach to large-scale machine\nlearning, helps with the computational feasibility, but at the cost of losing\nthe interpretability of features. In this paper, we present MISSION, a novel\nframework for ultra large-scale feature selection that performs stochastic\ngradient descent while maintaining an efficient representation of the features\nin memory using a Count-Sketch data structure. MISSION retains the simplicity\nof feature hashing without sacrificing the interpretability of the features\nwhile using only O(log^2(p)) working memory. We demonstrate that MISSION\naccurately and efficiently performs feature selection on real-world,\nlarge-scale datasets with billions of dimensions.","url_abs":"http://arxiv.org/abs/1806.04310v1","url_pdf":"http://arxiv.org/pdf/1806.04310v1.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":"mission-ultra-large-scale-feature-selection","repo_url":"https://github.com/rdspring1/MISSION","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"},{"task_slug":"feature-selection","task_name":"feature selection"}],"methods":[{"method_slug":"interpretability","method_name":"Interpretability"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1806.04310","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}