BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Latent-variable regression for spectroscopic measurements.
DTSTART:20090925T101500
DTSTAMP:20260407T084705Z
UID:230512b0abfcda25a57685c5c144b3c6e0ca5ec08a3a361a58b2fd44
CATEGORIES:Conferences - Seminars
DESCRIPTION:M. P. Gujral\, Laboratoire d'Automatique\, EPFL\nLatent-variab
 le regression has found use in many diverse applications such as social sc
 ience\, business studies and engineering. I will focus the attention on th
 e engineering application in spectroscopy (e.g. near-infrared\, infrared\,
  Raman\, chromatography\, nuclear magnetic resonance\, mass spectroscopy)\
 , where a typical goal is to estimate analyte concentrations from the corr
 esponding spectra. Spectroscopic data is characterized by (i) a high degre
 e of multicollinearity in the spectra\, and (ii) a linear relationship bet
 ween the spectra and the concentrations. These properties make latent-vari
 able regression models\, such as PCR and PLSR\, ideal candidates for calib
 ration. The talk will be tutorial in nature. In the first part\, I will in
 troduce PCR and PLSR\, the statistics (or heuristics) behind these methods
 \, and a cautionary note on some pitfalls or common misuses of these metho
 ds. In the second part\, I will present my thesis work on the correction o
 r update of PCR/PLSR models when affected by systematic disturbances.
LOCATION:MEC2405
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR