Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Alleviating tuning sensitivity in Approximate Dynamic Programming


P. Beuchat, A. Georghiou, J. Lygeros

European Control Conference (ECC), [OC:03715]

The simplicity of Approximate Dynamic Program- ming offers benefits for large-scale systems compared to other synthesis and control methodologies. A common technique to approximate the Dynamic Program, is through the solution of the corresponding Linear Program. The major drawback of this approach is that the online performance is very sensitive to the choice of tuning parameters, in particular the state relevance weighting parameter. Our work aims at alleviating this sensitivity. To achieve this, we propose to find a set of approximate Q-functions, each for a different choice of the tuning paramters, and then to use the point-wise maximum of the set of Q-functions for the online policy. The pointwise maximum promises to be better than using only one of individual Q-functions for the online policy. We demonstrate that this approach immunizes against tuning errors in the parameter selection process by application to a stylized portfolio optimization problem.

The code used to generate the results for the numerical examples can be found here: (url to be inserted)


Type of Publication:


No Files for download available.
% Autogenerated BibTeX entry
@InProceedings { BeuGeo:2016:IFA_5403,
    author={P. Beuchat and A. Georghiou and J. Lygeros},
    title={{Alleviating tuning sensitivity in Approximate Dynamic
    booktitle={European Control Conference (ECC)},
Permanent link