{"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/causality-networks","title":"Causality Networks","arxiv_id":"1406.6651","date":"2014-06-25","proceeding":null,"authors":["Ishanu Chattopadhyay"],"abstract":"While correlation measures are used to discern statistical relationships\nbetween observed variables in almost all branches of data-driven scientific\ninquiry, what we are really interested in is the existence of causal\ndependence. Designing an efficient causality test, that may be carried out in\nthe absence of restrictive pre-suppositions on the underlying dynamical\nstructure of the data at hand, is non-trivial. Nevertheless, ability to\ncomputationally infer statistical prima facie evidence of causal dependence may\nyield a far more discriminative tool for data analysis compared to the\ncalculation of simple correlations. In the present work, we present a new\nnon-parametric test of Granger causality for quantized or symbolic data streams\ngenerated by ergodic stationary sources. In contrast to state-of-art binary\ntests, our approach makes precise and computes the degree of causal dependence\nbetween data streams, without making any restrictive assumptions, linearity or\notherwise. Additionally, without any a priori imposition of specific dynamical\nstructure, we infer explicit generative models of causal cross-dependence,\nwhich may be then used for prediction. These explicit models are represented as\ngeneralized probabilistic automata, referred to crossed automata, and are shown\nto be sufficient to capture a fairly general class of causal dependence. The\nproposed algorithms are computationally efficient in the PAC sense; $i.e.$, we\nfind good models of cross-dependence with high probability, with polynomial\nrun-times and sample complexities. The theoretical results are applied to\nweekly search-frequency data from Google Trends API for a chosen set of\nsocially \"charged\" keywords. The causality network inferred from this dataset\nreveals, quite expectedly, the causal importance of certain keywords. It is\nalso illustrated that correlation analysis fails to gather such insight.","url_abs":"http://arxiv.org/abs/1406.6651v1","url_pdf":"http://arxiv.org/pdf/1406.6651v1.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":"causality-networks","repo_url":"https://github.com/zeroknowledgediscovery/Cynet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}