Optimization and Design of ExperimentS (ODES)
PACE Workshop
Organizers: I. Poli, M. Forlin
PURPOSE The purpose of the workshop is:
• to discuss Designs and Models of combinatorially complex experiments.
• to discuss optimization criteria for experimentation.
• to develop future collaborations between different groups of theoreticians as well as with
some experimental groups.
• to outline an agenda of key topics to be addressed in future experimental design activities
PROGRAM Friday 15th June09:00 – 09:45 Reception
09:45 – 10:00 Opening Session- Welcome
Norman Packard , CoDirector, ECLT
10:00 – 10:45 Modeling Evolutionary Designs of Experiments
Irene Poli, University Ca’ Foscari, Venice, Italy
10:45 Coffee break
11:15 – 12:00 Genetic Algorithms as Screening Designs
James Cawse, ProtoLife, Venice, Italy
12:00 – 12:45 Model-based Genetic Algorithms for Designing
Mixture Experiments
Michele Forlin, University Ca’ Foscari, Venice, Italy
13:00 Lunch (Palazzo Franchetti Cafeteria)
14:30 – 15:15 Practical experience of integrating Genetic
Programming and statistical modeling
Flor Castillo, The Dow Chemical Company, Freeport, TX, USA
15:15 – 16:00 Evolutionary Algorithms for Self-Assembling
Systems: Simulations as a Tool for the Development of Novel Optimization Strategies
Rudolf Füchslin, Ruhr University Bochum, Germany
16:00 Coffee break
16:30 – 17:15 Nonlinearity Reduction and Ordinal Optimization
for modeling hard problems in industrial data analysis (New flavours in Genetic Programming)
Katya Vladislavleva, Department of Operations Research, Tilburg
University, The Netherlands
17:15 – 18:00 Evolutionary computing: what are the differences?
Richard Walker, Xiwrite S.a.s., Rome, Italy
Saturday, June 16 09:15 – 10:00 A Gestalt Genetic Algorithm: Less details for better
search
Hugues Bersini, IRIDIA, Université Libre de Bruxelles, Brussels,
Belgium
10:00 – 10:45
Microfluidics in an integrated system: pros and
cons of the pico-liter chemistry
Uwe Tangen, Ruhr University Bochum, Germany
10:45 Coffee break
11:15 – 12:00 Variable Selection in industrial datasets using
Pareto Symbolic regression via Genetic
Programming
Guido Smits, Department of Engineering Sciences, Dow Chemical
Company, Belgium
12:00 – 12:45 D-optimal designs for estimating of extremum
point of multivariate quadratic regression model
Andrey Pepelyshev, Department of Mathematics, Saint-Petersburg State University, Russia
13:00 Lunch (Palazzo Franchetti Cafeteria)