BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Data-Driven Engineering Design: Data-assisted high-fidelity modeli
 ng for systems design and monitoring
DTSTART:20190213T124500
DTEND:20190213T134500
DTSTAMP:20260407T002712Z
UID:02f4fa5db0409d7cca40daac35646b20d2297287523089ecd7a4af37
CATEGORIES:Conferences - Seminars
DESCRIPTION:Audrey Olivier\, Hopkins Extreme Materials Institute\,\nJohns 
 Hopkins University\, Baltimore\, MD\, USA\n\n \nIncreased availability of
  large amounts of data\, development of new sensing technologies and conne
 cted systems have generated tremendous interest in the development of meth
 ods to learn from data. In parallel\, engineers have a long history of bui
 lding high-fidelity models\, based on physical principles\, that allow us 
 to model the behavior of highly complex systems. This talk aims at present
 ing some of the exciting research opportunities that arise at the intersec
 tion of these two topics\, data analytics\, and high-fidelity modeling\, a
 nd illustrate potential applications. \n\nFor instance\, system identific
 ation methods make use of measurements from a system to learn the equation
 s and parameters characterizing it\, possibly in real-time. Such learning 
 methods find a wide variety of applications\, such as calibration and vali
 dation of physics-based models of systems during their design process\, or
  diagnostics of mechanical systems. Bayesian inference algorithms are part
 icularly attractive in this framework as they allow quantification of unce
 rtainties around the estimated quantities of interest\, where large uncert
 ainties can arise when the data is not highly informative\, or the inputs 
 or other experimental variables are uncertain. However\, Bayesian techniqu
 es become highly computationally expensive when used for inference in larg
 e dimensional nonlinear systems\, for instance\, systems represented by fi
 nite element models. This talk aims at demonstrating the potential of Baye
 sian filtering techniques\, such as nonlinear Kalman filters or particle f
 ilters\, with respect to parameter estimation\, and present algorithmic en
 hancements that achieve a good trade-off between accuracy and computationa
 l cost. Furthermore\, we will show how these probabilistic learning proced
 ures can also be integrated into more complex frameworks\, such as model s
 election and optimal design of experiments.\n\nThe combination of data-min
 ing and physics-based modeling also finds applications in various engineer
 ing fields. In the materials sciences\, for instance\, machine learning al
 gorithms can be used in conjunction with experiments or high-fidelity simu
 lations to obtain a better understanding of structure-property-performance
  linkages\, enabling efficient microstructure characterization and inverse
  design of materials. Very interestingly\, the application of machine lear
 ning algorithms to engineering problematics gives rise to challenging prob
 lematics and technical questions\, for example\, the need to accurately qu
 antify uncertainties through machine learning pipelines\, or handling smal
 l amounts of data. This talk thus aims at illustrating some of the challen
 ges and opportunities related to the use of both model-based and data-base
 d learning algorithms for engineering applications. \n\nBio of speaker: A
 udrey Olivier pursued her undergraduate studies in her native France befor
 e moving to Columbia University\, New York\, for graduate studies. She hol
 ds a Diplôme d’Ingénieur (graduate degree) from École Centrale de Nan
 tes\, France\, and a Master’s of Science and Ph.D. from the Department o
 f Civil Engineering and Engineering Mechanics at Columbia University. Her 
 doctoral research primarily focused on the development of efficient system
  identification and uncertainty quantification tools\, such as Bayesian in
 ference algorithms\, for inverse modeling of large dimensional complex sys
 tems. Dr. Olivier is currently appointed as a Postdoctoral Fellow in the H
 opkins Extreme Materials Institute at Johns Hopkins University. She is wor
 king on the development of machine learning algorithms for materials model
 ing. Dr. Olivier received a Teaching Assistant Excellence Award from her D
 epartment at Columbia University in 2015\, and a Mindlin Scholar in Civil 
 Engineering and Engineering Mechanics award in 2017 for her work as a doct
 oral student. Dr. Olivier is also involved in outreach programs aiming at 
 introducing young people to STEM fields.
LOCATION:MEB1 B10 https://plan.epfl.ch/?room==ME%20B1%20B10
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR