BEGIN:VCALENDAR
VERSION:2.0
METHOD:PUBLISH
PRODID:-//Telerik Inc.//Sitefinity CMS 12.2//EN
BEGIN:VTIMEZONE
TZID:Mountain Standard Time
BEGIN:STANDARD
DTSTART:20201102T020000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYHOUR=2;BYMINUTE=0;BYMONTH=11
TZNAME:Mountain Standard Time
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20200301T020000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYHOUR=2;BYMINUTE=0;BYMONTH=3
TZNAME:Mountain Daylight Time
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
DESCRIPTION:\nZoom Meeting ID: 950 0967 8295\nPasscode: \;591441\nSpeak
er: Erin Austin\, PhD\nAssistant Professor\, Mathematical and Statistical
Sciences\, UCDenver\nTitle: \;Using Data Consistent Inversion to descr
ibe and quantify the impact of sources of uncertainty on predictions of lu
ng function in the COPDGene cohort \;\n\nAbstract: \;Data Consiste
nt Inversion (DCI) is a new iterative methodology within the field of unce
rtainty quantification to identify distributional forms for a model&rsquo\
;s parameters that lead to predictions consistent with observed outcomes.
That is\, DCI is a new way to quantify if a hypothesized model can be used
to generate predictions that are distributionally similar to that of the
observed outcomes. The method has traditionally been limited to areas wher
e hard science has credibly determined the functional form (model) of a be
havior\, and uncertainty quantification is often used to study small chang
es to this known model. In our work\, we show that statistical reasoning c
an substitute for the role of hard science and allow us to credibly hypoth
esize models for our phenomenon of interest. We can then adapt DCI tools t
o provide researchers a new metric to assess whether our hypothesized mode
ls give data consistent predictions. Further\, the flexibility of DCI perm
its us to easily consider modifications to linear models\; for example\, e
mbed unobserved measurement error in observed variables. The process resul
ts in a model form\, including distributions for all its parameters\, that
has been optimally adapted for consistency with our data\; a model that c
an then be used to better quantify the effect of the different model compo
nents on prediction.\nSpecifically\, we use DCI to study different hypothe
sized model forms for predicting 10-year measurements of lung health from
baseline and 5-year measurements for participants in the COPDGene project.
Using DCI we demonstrate how a linear model that accounts for bias in pre
vious measurements is sufficient for predictions. Further\, we show that c
lassic regression approaches underestimate the true range in 10-years outc
omes as they fail to adequately capture the true uncertainty in the data.
DTEND:20210407T190000Z
DTSTAMP:20210415T040108Z
DTSTART:20210407T180000Z
LOCATION:
SEQUENCE:0
SUMMARY:BIOS Seminar Series: Erin Austin\, PhD
UID:RFCALITEM637540344689169770
X-ALT-DESC;FMTTYPE=text/html:\n**Zoom Meeting ID:
**950 0967 8295

\n**Pas
scode: \;**591441

\n**Speaker: **Erin Austin\, PhD

\nAssistant Professor\, Mathemati
cal and Statistical Sciences\, UCDenver

\n**Title: \;Using Data Consistent Inversion to describe and quantify the impact of
sources of uncertainty on predictions of lung function in the COPDGene coh
ort \;**

\n\n**Abstract: \;**Data
Consistent Inversion (DCI) is a new iterative methodology within the fiel
d of uncertainty quantification to identify distributional forms for a mod
el&rsquo\;s parameters that lead to predictions consistent with observed o
utcomes. That is\, DCI is a new way to quantify if a hypothesized model ca
n be used to generate predictions that are distributionally similar to tha
t of the observed outcomes. The method has traditionally been limited to a
reas where hard science has credibly determined the functional form (model
) of a behavior\, and uncertainty quantification is often used to study sm
all changes to this known model. In our work\, we show that statistical re
asoning can substitute for the role of hard science and allow us to credib
ly hypothesize models for our phenomenon of interest. We can then adapt DC
I tools to provide researchers a new metric to assess whether our hypothes
ized models give data consistent predictions. Further\, the flexibility of
DCI permits us to easily consider modifications to linear models\; for ex
ample\, embed unobserved measurement error in observed variables. The proc
ess results in a model form\, including distributions for all its paramete
rs\, that has been optimally adapted for consistency with our data\; a mod
el that can then be used to better quantify the effect of the different mo
del components on prediction.

\nSpecifically\, we use DCI to study di
fferent hypothesized model forms for predicting 10-year measurements of lu
ng health from baseline and 5-year measurements for participants in the CO
PDGene project. Using DCI we demonstrate how a linear model that accounts
for bias in previous measurements is sufficient for predictions. Further\,
we show that classic regression approaches underestimate the true range i
n 10-years outcomes as they fail to adequately capture the true uncertaint
y in the data.

END:VEVENT
END:VCALENDAR