Homepage of Carola Doerr (formerly Winzen)

photo of Carola Doerr

Carola Doerr
Sorbonne University
(formerly Université Pierre et Marie Curie - Paris 6)
LIP6, équipe Recherche Opérationnelle
case 169, 4 place Jussieu
75252 Paris Cedex 05
France

Email: Carola [dot] Doerr "at" mpi-inf.mpg.de
Phone: 0033 (0)1 44 27 54 42


Some Selected Activities and News


  • [New article] A short article about nature-inspired optimization and the role of theory is available (in French) at Images des mathématiques. I will try to make an English version available soon. Special thanks to Romain Dujardin for suggesting to write this article and for translating it!
  • IOHprofiler: Together with Hao Wang, Furong Ye, Sander van Rijn, and Thomas Bäck from Leiden University, we have created IOHprofiler, an algorithm benchmarking and profiling tool for iterative optimization heuristics.
    IOHprofiler on GitHub, here is a (preliminary) documentation, and here you can sign up for the newsletter to receive important updates about IOHprofiler.
  • [Projects, Internships] Two new projects just got accepted. Students interested in working with me as research interns (e.g., ``M1/M2 stages'') are invited to get in touch. Potential internship topics include:
    - (online) algorithm configuration (e.g., can we leverage techniques from machine learning to select parameters ``on the fly''?),
    - algorithm benchmarking (which problems to include in a benchmark set, how to measure performance, ...),
    - computational aspects of discrepancy measures.
    I am interested in both theoretical and empirical aspects of these problems.
    Here is an example for an internship project: Automated Algorithm Configuration - feel free to reach out if you want to know more about this or the above-mentioned projects!
  • Together with Dirk Arnold I am chairing the program committee of FOGA 2019, which takes place in Potsdam, Germany, from August 26 to 29.
  • 2018 tutorial slides on dynamic parameter choices in evolutionary computation are available here: DoerrGECCO18tutorial.pdf. The target audience of this tutorial are researchers familiar with evolutionary computation. I am happy to discuss all concepts and ideas in a form that avoids community-specific terminology.
  • Two new survey articles available:
    C. Doerr: Complexity Theory for Black-Box Optimization Heuristics (link to arXiv version) and
    B. Doerr and C. Doerr: Theory of Parameter Control for Discrete Black-Box Optimization: Provable Performance Gains Through Dynamic Parameter Choices (link to arXiv version)
  • Organizing Dagstuhl seminar 19431 on Theory of Randomized Optimization Heuristics (October 20-25, 2019), together with Carlos M. Fonseca, Tobias Friedrich, and Xin Yao.
  • Our university has changed its name from Université Pierre et Marie Curie (Paris 6) to Sorbonne Université.
  • Dirk Sudholt and I have been chairing THEORY track of GECCO 2017. We are now editing a special issue in Algorithmica.
  • I am serving as Vice-chair of COST action 15140 on Improving Applicability of Nature-Inspired Optimisation by Joining Theory and Practice (ImAppNIO), and have also organized our first training school COST training school here in Paris (7 days, 35 participants, 7 teachers).

Research Interests


  • Mathematical and empirical aspects of iterative optimization heuristics (aka randomized search heuristics) such as local search algorithms, evolutionary algorithms, and other black-box optimization techniques
    At the moment, I am particularly interested in online algorithm configuration and different aspects of benchmarking.
  • Black-box complexity, aka randomized query complexity
  • Randomized and quasi-randomized algorithms in general
  • Geometric discrepancies, in particular computational aspects of the star discrepancy

Publications


A list of publications can be found here. Here is an outdated (not sure when I will be able to update...) list of my talks.


Teaching


Courses:

PhD Students:

  • Advisor of PhD student Anja Jankovic (10/2018-).
    Subject: Online Algorithm Configuration
  • Co-Advisor of PhD student Jing Yang (10/2015-09/18).
    Title of her PhD thesis: From a Complexity Theory of Evolutionary Computation to Superior Randomized Search Heuristics

Bachelor/Master Students and PhD interns:

  • Advisor of student intern Nathan Buskulic (summer 2018).
    Title of the project: Optimal Evolutionary Algorithms with Dynamic Parameters
  • Co-Advisor of Master student Anja Jankovic (summer 2018).
    Title of the project: Randomness in Scheduling
  • Advisor of Master student Eduardo Carvalho Pinto (summer 2017).
    Title of the project: Self-Adjusting Parameter Choices for Discrete Black-Box Optimization
  • Co-Advisor of Master student Jing Yang (summer 2015).
    Title of the project: Tight Bounds for the Unbiased Black-Box Complexity of OneMax
  • Co-Advisor of Master student Axel de Perthuis de Laillevault (summer 2014).
    Title of the project: Evolutionary Algorithms with Iterated Initial Sampling
  • Co-Advisor of the Master student Franziska Ebel (defended spring 2013).
    Title of the thesis: Lessons from the Black-Box: Fast Crossover-Based Genetic Algorithms
  • Advisor of PhD student intern G. Ramakrishna (summer 2012).
    Title of the project: Computing Minimum Cycle Bases in Graphs of Bounded Treewidth
  • Advisor of Master student Vijay Ingalalli (defended autumn 2011).
    Title of the thesis: Evolutionary Algorithms to Compute Lower Bounds for the Star Discrepancy
  • Co-Advisor of student intern Jong-Hyun Lee (winter 2011/12).
    Title of the project: Playing Mastermind with Constant Size Memory

Academic Activities



Short CV


Page last modified October 12, 2018.