{"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/learning-independent-causal-mechanisms","title":"Learning Independent Causal Mechanisms","arxiv_id":"1712.00961","date":"2017-12-04","proceeding":"ICML 2018 7","authors":["Giambattista Parascandolo","Niki Kilbertus","Mateo Rojas-Carulla","Bernhard Schölkopf"],"abstract":"Statistical learning relies upon data sampled from a distribution, and we\nusually do not care what actually generated it in the first place. From the\npoint of view of causal modeling, the structure of each distribution is induced\nby physical mechanisms that give rise to dependences between observables.\nMechanisms, however, can be meaningful autonomous modules of generative models\nthat make sense beyond a particular entailed data distribution, lending\nthemselves to transfer between problems. We develop an algorithm to recover a\nset of independent (inverse) mechanisms from a set of transformed data points.\nThe approach is unsupervised and based on a set of experts that compete for\ndata generated by the mechanisms, driving specialization. We analyze the\nproposed method in a series of experiments on image data. Each expert learns to\nmap a subset of the transformed data back to a reference distribution. The\nlearned mechanisms generalize to novel domains. We discuss implications for\ntransfer learning and links to recent trends in generative modeling.","url_abs":"http://arxiv.org/abs/1712.00961v5","url_pdf":"http://arxiv.org/pdf/1712.00961v5.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":"learning-independent-causal-mechanisms","repo_url":"https://github.com/kevtimova/licms","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1712.00961","atlas_url":"https://app.syntology.ai/?focus=1712.00961","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}