BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Screening and optimization with one single OMARS design: design\, 
 analysis\, and optimization
DTSTART:20240915T120000Z
DTEND:20240915T160000Z
DTSTAMP:20260414T114700Z
UID:indico-event-58@conferences.enbis.org
DESCRIPTION:Speakers: Peter Goos (KU Leuven)\n\n\nScreening and optimizati
 on with one single OMARS design: design\, analysis\, and optimization\nPar
 t of the ENBIS-24 Leuven conference.  \nAfter the course\, there will be 
 a meet and greet with Peter Goos and Bradley Jones at 18h\, and a drink of
 fered by Effex at 19h. \nInstructor\nJosé Núñez Ares\, EFFEX\nOverview
 \nThe standard response surface methodology of conducting a screening expe
 riment followed by an optimisation experiment has been challenged in recen
 t years. In 2011\, the emergence of a new type of design\, the definitive 
 screening design (DSD) (Jones & Nachstheim\, 2011)\, posed a new challenge
  to this approach. In 2020\, the first paper on OMARS designs (Núñez Are
 s & Goos\, 2020) appeared\, generalising DSDs and extending the use of a s
 ingle design for screening and optimisation due to its orthogonality prope
 rties. \nThe selection of the best experimental design among different alt
 ernatives is crucial\, on the one hand to minimise the experimental effort
  and\, on the other hand\, to maximise the information obtained once the e
 xperiment has been performed and the data collected. To achieve these goal
 s\, it is important to balance the size and different quality characterist
 ics of the designs\, such as projection estimation capacity\, the power to
  detect different effects\, or the number of replicate points. \n The ana
 lysis of screening + optimisation experiments involves a large number of f
 actors\, making the analysis of experimental data difficult. Recently\, a 
 novel algorithm for all-subset model selection has emerged that can cope w
 ith problems with more than 100 potential effects and has been successfull
 y applied to industrial problems (Vázquez\, Schoen & Goos\, 2020). \nOpti
 misation of multi-response problems is often the end goal. The trade-off b
 etween the different responses and the high dimensionality of the input sp
 ace (high number of factors) makes it challenging. The probability of succ
 ess of being within specifications uses the predictive power of the underl
 ying statistical models and quantifies the uncertainty and robustness of a
 ny given combination of factor values.\nIn this short course\, we will gui
 de the user through this process using our powerful yet intuitive design s
 election tool\, new model selection algorithm and the optimization platfor
 m. This way\, the best possible designs are coupled to the best possible m
 odels to analyze the precious data resulting from them.\nOutline\nThe cour
 se is divided in two parts:\nPart 1: how to choose an experimental design 
 for screening + optimization\n\nScreening\, optimization\, or screening + 
 optimization simultaneously?\nQuality attributes of a design and study of 
 their trade-offs. \nCase studies.\n\n \nPart 2: how to model experimental
  data involving a large number of factors and how to optimize multiple res
 ponses simultaneously\n\nAll-subset model selection for problems with a hi
 gh input dimensionality\nHow to compare different models with each other\n
 Multi-response optimization problems and calculation of the probability of
  success\nCase studies\n\nPacticalities\n\nEach participant will receive a
  handout with the slides that will be used during the workshop\nParticipan
 ts are invited to bring their own laptop. A trial account of the EFFEX sof
 tware will be provided. Since it is cloud-based\, no installation is requi
 red\nThe instructor will guide the participant through all practical cases
 \, in such a way that the participants will be able to replicate what the 
 instructor does\n\nMeet & greet with Peter Goos and Bradley Jones \nAfter
  the course\, there is a meet and greet with Peter Goos and Bradley Jones.
  \nDrink in the city center \nParticipants are cordially invited for a dr
 ink at Leuven Centraal\, Margarethaplein 3\, 3000 Leuven. It is a 10 minut
 e walk from the course venue.\nDrinks start at 19h!\nShort bio\n \n\nJos
 é Núñez Ares was a postdoctoral researcher at KU Leuven's MeBioS resear
 ch group (Mechatronics\, Biostatistics and Sensors)\, where he studied way
 s to set up new\, cost-efficient experimental designs.\nHe obtained a Mast
 er in Operations Research from Erasmus University in Rotterdam and a Bache
 lor in Civil Engineering from the University of Coruña (Spain). He obtain
 ed his PhD at KU Leuven under the supervision of Prof Peter Goos.\nHis res
 earch has been published in top journals and he is known in the academic c
 ommunity as the inventor\, together with Prof Goos\, of the OMARS design m
 ethodology. Besides his research\, José is active in industrial consultan
 cy and has successfully delivered integrated DoE solutions to companies in
  the pharmaceutical\, chemical\, energy and manufacturing sectors. He hold
 s a US patent on an algorithm for experimental design selection.\nJosé N
 úñez Ares is co-founder of EFFEX and assumes the role of Chief Scientifi
 c Officer.\nwww.effex.app \n \n\nhttps://conferences.enbis.org/event/58/
LOCATION:Conference Room 3 (Irish College Leuven)
URL:https://conferences.enbis.org/event/58/
END:VEVENT
END:VCALENDAR