Return to the list
post-doc: Transfer Learning for Evolutionary Algorithms Applied to Optimization
Where: Amsterdam,
Netherlands
What: CWI is hosting a 2.5-year project focusing on Continuous Tailoring of Evolutionary Algorithms to Recurring Problems. We are looking for a talented postdoc who is interested in pursuing scientific research at the intersection of evolutionary algorithms and machine learning.
The project targets settings, where similar optimisation problems are encountered repeatedly. For example, logistics companies need to solve comparable routing problems every day. Medical professionals need to create effective treatment plans regularly. Hyperparameters of algorithms need to be adjusted for each new task. While the underlying problem structures remain unknown, black-box optimisation approaches could be iteratively tailored to the regularly encountered problem over time.
The main question we want to answer is: Which steps of the tailoring process of evolutionary algorithms can be transferred across which different usecases? Related research topics thus include measuring problem similarity, as well as analysing and explaining algorithm behaviour (theoretically and empirically).
Link: https://www.cwi.nl/en/jobs/vacancies/1126083/
Who: Vanessa Volz, vanessa.volz@cwi.nl
When: Until 2025-01-09 18:00
Presented at next GECCO?: yes
Return to the list
post-doc: Transfer Learning for Evolutionary Algorithms Applied to Optimization
Where: Amsterdam,
NetherlandsWhat: CWI is hosting a 2.5-year project focusing on Continuous Tailoring of Evolutionary Algorithms to Recurring Problems. We are looking for a talented postdoc who is interested in pursuing scientific research at the intersection of evolutionary algorithms and machine learning.
The project targets settings, where similar optimisation problems are encountered repeatedly. For example, logistics companies need to solve comparable routing problems every day. Medical professionals need to create effective treatment plans regularly. Hyperparameters of algorithms need to be adjusted for each new task. While the underlying problem structures remain unknown, black-box optimisation approaches could be iteratively tailored to the regularly encountered problem over time.
The main question we want to answer is: Which steps of the tailoring process of evolutionary algorithms can be transferred across which different usecases? Related research topics thus include measuring problem similarity, as well as analysing and explaining algorithm behaviour (theoretically and empirically).
Link: https://www.cwi.nl/en/jobs/vacancies/1126083/
Who: Vanessa Volz, vanessa.volz@cwi.nl
When: Until 2025-01-09 18:00
Presented at next GECCO?: yes
Return to the list