Institute Homepage DE Sign In

Back

Empirical Inference Conference Paper 2008

Fitness Expectation Maximization

Thumb ticker sm 12009745 10103538825457245 7502907506146263960 n
Empirical Inference
Jan Peters
Research Group Leader

We present Fitness Expectation Maximization (FEM), a novel method for performing ‘black box’ function optimization. FEM searches the fitness landscape of an objective function using an instantiation of the well-known Expectation Maximization algorithm, producing search points to match the sample distribution weighted according to higher expected fitness. FEM updates both candidate solution parameters and the search policy, which is represented as a multinormal distribution. Inheriting EM’s stability and strong guarantees, the method is both elegant and competitive with some of the best heuristic search methods in the field, and performs well on a number of unimodal and multimodal benchmark tasks. To illustrate the potential practical applications of the approach, we also show experiments on finding the parameters for a controller of the challenging non-Markovian double pole balancing task.

Author(s): Wierstra, D. and Schaul, T. and Peters, J. and Schmidhuber, J.
Book Title: PPSN 2008
Journal: Parallel Problem Solving from Nature – PPSN X
Pages: 337-346
Year: 2008
Month: September
Day: 0
Editors: Rudolph, G. , T. Jansen, S. Lucas, C. Poloni, N. Beume
Publisher: Springer
Bibtex Type: Conference Paper (inproceedings)
Address: Berlin, Germany
DOI: 10.1007/978-3-540-87700-4_34
Event Name: 10th International Conference on Parallel Problem Solving From Nature
Event Place: Dortmund, Germany
Digital: 0
Electronic Archiving: grant_archive
Language: en
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik
Links:

BibTex 

@inproceedings{6888,
  title = {Fitness Expectation Maximization},
  journal = {Parallel Problem Solving from Nature – PPSN X},
  booktitle = {PPSN 2008},
  abstract = {We present Fitness Expectation Maximization (FEM), a novel method for performing ‘black box’ function optimization. FEM searches the fitness landscape of an objective function using an instantiation of the well-known Expectation Maximization algorithm, producing search points to match the sample distribution weighted according to higher expected fitness. FEM updates both candidate solution parameters and the search policy, which is represented as a multinormal distribution. Inheriting EM’s stability and strong guarantees, the method is both elegant and competitive with some of the best heuristic search methods in the field, and performs well on a number of unimodal and multimodal benchmark tasks. To illustrate the potential practical applications of the approach, we also show experiments on finding the parameters for a controller of the challenging non-Markovian double pole balancing task.},
  pages = {337-346},
  editors = {Rudolph, G. , T. Jansen, S. Lucas, C. Poloni, N. Beume},
  publisher = {Springer},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {Berlin, Germany},
  month = sep,
  year = {2008},
  slug = {6888},
  author = {Wierstra, D. and Schaul, T. and Peters, J. and Schmidhuber, J.},
  month_numeric = {9}
}