BEGIN:VCALENDAR VERSION:2.0 PRODID:researchseminars.org CALSCALE:GREGORIAN X-WR-CALNAME:researchseminars.org BEGIN:VEVENT SUMMARY:Milan Studeny (Institute of Information Theory and Automation\, Pr ague) DTSTART:20230301T160000Z DTEND:20230301T171500Z DTSTAMP:20260423T021357Z UID:AAIT/6 DESCRIPTION:Title: On conditional Ingleton inequalities\nby Milan Studeny (Institute of Info rmation Theory and Automation\, Prague) as part of Seminar on Algorithmic Aspects of Information Theory\n\n\nAbstract\nLinear information inequaliti es valid for entropy functions induced by discrete random variables play a n important role in the task to characterize discrete conditional independ ence structures. Specifically\, the so-called conditional Ingleton inequal ities in the case of 4 random variables are in the center of interest: the se are valid under conditional independence assumptions on the inducing ra ndom variables. The four inequalities of this form were earlier revealed: by Yeung and Zhang (1997)\, by Matúš (1999) and by Kaced and Romashchenk o (2013). In our recent paper (2021) the fifth inequality of this type was found. These five information inequalities can be used to characterize al l conditional independence structures induced by four discrete random vari ables. One of open problems in that 2021 paper was whether the list of con ditional Ingleton inequalities over 4 random variables is complete: the an alysis can be completed by a recent finding of Boege (2022) answering that question.\n LOCATION:https://researchseminars.org/talk/AAIT/6/ END:VEVENT END:VCALENDAR