BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal iCal API//EN
X-WR-CALNAME:Events items teaser
X-WR-TIMEZONE:America/Toronto
BEGIN:VTIMEZONE
TZID:America/Toronto
X-LIC-LOCATION:America/Toronto
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20241103T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:69d9342362e90
DTSTART;TZID=America/Toronto:20250923T103000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20250923T113000
URL:https://uwaterloo.ca/statistics-and-actuarial-science/events/joint-dist
 inguished-lecture-bin-yu
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West Room: DC 1302 Waterloo ON N2L 3G1 Canada
SUMMARY:Joint Distinguished Lecture by Bin Yu
CLASS:PUBLIC
DESCRIPTION:Dean's Distinguished Women in Mathematics\, Statistics and Comp
 uter\nScience Lecture Series & David Sprott Distinguished Lecture Series\n
 \nBIN YU\n_CDSS Chancellor's Distinguished Professor\, Statistics\, EECS\,
  Center\nfor Computational Biology\nSenior Advisor\, Simons Inst for the T
 heory of Computing\nMember\, U.S. National Academy of Sciences\, 2014\nMem
 ber\, American Academy of Arts and Sciences\, 2013\nGuggenheim Fellow\, 20
 06_\n\nRoom: DC 1302\n\n-------------------------\n\nVERIDICAL DATA SCIENC
 E TOWARDS TRUSTWORTHY AI \n\nIn this talk\, I will introduce the\nPredict
 ability-Computability-Stability (PCS) framework for veridical\n(truthful)
  data science\, highlighting its critical role in producing\nreliable and
  actionable insights. I will share success stories\nfrom cancer detection
  and cardiology\, showcasing how PCS principles\nhave guided cost effecti
 ve designs and improved outcomes in these\nprojects. Since trustworthy unc
 ertainty quantification is\nindispensable for trustworthy AI\, I will disc
 uss PCS uncertainty\nquantification for prediction in regression and mult
 i-class\nclassification. PCS-UQ consists of three steps: pred-check\,\nbo
 otstrap\, and multiplicative calibration. Through test results over\n26 be
 nchmark datasets\, PCS-UQ will be shown to outperform common\nforms of co
 nformal prediction in terms of width\, subgroup coverage\,\nand subgroup 
 interval width. Finally\, the multiplicative step in\nPCS-UQ will be shown
  to be a new form of conformal prediction.
DTSTAMP:20260410T173219Z
END:VEVENT
END:VCALENDAR