{"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/hierarchical-methods-of-moments","title":"Hierarchical Methods of Moments","arxiv_id":"1810.07468","date":"2018-10-17","proceeding":"NeurIPS 2017 12","authors":["Matteo Ruffini","Guillaume Rabusseau","Borja Balle"],"abstract":"Spectral methods of moments provide a powerful tool for learning the\nparameters of latent variable models. Despite their theoretical appeal, the\napplicability of these methods to real data is still limited due to a lack of\nrobustness to model misspecification. In this paper we present a hierarchical\napproach to methods of moments to circumvent such limitations. Our method is\nbased on replacing the tensor decomposition step used in previous algorithms\nwith approximate joint diagonalization. Experiments on topic modeling show that\nour method outperforms previous tensor decomposition methods in terms of speed\nand model quality.","url_abs":"http://arxiv.org/abs/1810.07468v1","url_pdf":"http://arxiv.org/pdf/1810.07468v1.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":"hierarchical-methods-of-moments","repo_url":"https://github.com/mruffini/Hierarchical-Methods-of-Moments","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"tensor-decomposition","task_name":"Tensor Decomposition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}