Evolutionary & Swarm Design
Namrata Khemka, Ph.D. Student
Address:
Artificial Intelligence Lab
Dept. of Computer Science,
University of Calgary,
Calgary, Alberta, Canada, T2N1N4
Calgary, Alberta, Canada, T2N1N4
Email:
Website:
Khemka, Namrata; Jacob, Christian
Exploratory Toolkit for Evolutionary And Swarm-based Optimization
IMS'06, Proc. Eight International Mathematica Symposium
At: Avignon, France, 2006.
Exploratory Toolkit for Evolutionary And Swarm-based Optimization
IMS'06, Proc. Eight International Mathematica Symposium
At: Avignon, France, 2006.
Khemka, Namrata
Comparing Particle Swarms and Evolution Strategies: Benchmarks & Application
MSc. Thesis, University of Calgary, 2005.
Comparing Particle Swarms and Evolution Strategies: Benchmarks & Application
MSc. Thesis, University of Calgary, 2005.
Khemka, Namrata; Jacob, Christian; Cole, Gerald
Making soccer kicks better: A study in Particle Swarm Optimization
and Evolution Strategies.
In: David Corne (Ed.) : IEEE Congress on Evolutionary Computation
At: Edinburgh, UK, 2005.
Making soccer kicks better: A study in Particle Swarm Optimization
and Evolution Strategies.
In: David Corne (Ed.) : IEEE Congress on Evolutionary Computation
At: Edinburgh, UK, 2005.
Graduate Student Workshop:
Khemka, Namrata; Jacob, Christian; Cole Gerald
Making soccer kicks better: A study in Particle Swarm Optimization
Genetic and Evolutionary Computation Conference
At: Washington DC, US, 2005.
Khemka, Namrata; Jacob, Christian; Cole Gerald
Making soccer kicks better: A study in Particle Swarm Optimization
Genetic and Evolutionary Computation Conference
At: Washington DC, US, 2005.
Jacob, Christian; Khemka, Namrata
Particle Swarm Optimization in Mathematica An Exploration Kit for
Evolutionary Optimization
IMS'04, Proc. Sixth International Mathematica Symposium
At: Banff, Canada, 2004.
Particle Swarm Optimization in Mathematica An Exploration Kit for
Evolutionary Optimization
IMS'04, Proc. Sixth International Mathematica Symposium
At: Banff, Canada, 2004.
Publications:
Wouldn’t it be nice to have a car that uses 1 liter of fuel per 100 kilometers? There is a constant effort to find something better than what currently exists, such as lowering the fuel consumption of cars. A significant number of important applications exist that can be optimized in economics, management, engineering, and life sciences. Clearly, it is beneficial for society to continually improve the best available alternatives i.e., to constantly optimize with respect to specific constraints.
For my master’s thesis I worked on a soccer kick optimization problem, where the goal was to find ‘optimized’ settings of control parameters for a kinematical model of 17 leg muscles such that a kicked ball travels as far and as fast as possible. Determining better and better solutions is a key challenge as the model includes 56 parameters, and these values yield multiple solutions, thus making it difficult to anticipate the influence of each different parameter. I had extended a local hill-climbing strategy by applying an evolutionary particle swarm optimization algorithm. This new algorithm yielded very good results. In the course of these investigations and experiments, I realized the importance of parallelization in order to solve large-scale optimization problems. This laid the groundwork for my proposed Ph.D. research to create a toolkit for parallel and distributed optimization of highly parameterized computer models.
