BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Economics Research - ECPv6.0.5//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Economics Research
X-ORIGINAL-URL:https://research.reading.ac.uk/economics
X-WR-CALDESC:Events for Economics Research
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:20230326T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:20231029T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/London:20231122T143000
DTEND;TZID=Europe/London:20231122T160000
DTSTAMP:20260421T191433
CREATED:20231106T184138Z
LAST-MODIFIED:20231114T125342Z
UID:3646-1700663400-1700668800@research.reading.ac.uk
SUMMARY:External Seminar by William Pouliot (University of Birmingham)
DESCRIPTION:Title: On kernel regression with NMAR response values\n \n\nAbstract: We consider the problem of kernel regression estimation in the presence of Not-Missing-At-Random (NMAR)\, or nonignorable\, response variables. Our  proposed approach involves two steps: In the first step\, we construct a family of models (possibly infinite dimensional) indexed by the unknown nonignorability component of the missing probability mechanism.  In the second step\, a search is carried out to find the empirically optimal member of an appropriate cover (or subclass) of  the underlying family in the sense of minimizing the mean squared prediction error. Our methods use a data-splitting  approach which is quite easy to implement.  We also derive exponential bounds on the performance of the resulting estimators in terms of their deviations from the true regression curve in general L_p norms\, where we also allow the size of the cover or subclass to diverge as the sample size n increases. These bounds together with the Borel-Cantelli lemma immediately yield various strong convergence results for the proposed estimators. As an application of our findings\, we consider the problem of nonparametric statistical classification based on the proposed regression estimators and also look into their rates of convergence in different settings. \n\nco-authors: M. Mojirsheibani and A. Shakhbandaryan\n\n \nYou can check out William’s latest research here: https://www.birmingham.ac.uk/staff/profiles/business/pouliot-william.aspx
URL:https://research.reading.ac.uk/economics/event/external-seminar-by-william-pouliot-university-of-birmingham/
LOCATION:Edith Morley 126\, United Kingdom
CATEGORIES:External Seminars
ORGANIZER;CN="Carl%20Singleton":MAILTO:c.a.singleton@reading.ac.uk
END:VEVENT
END:VCALENDAR