Some optimization problems, such as Molecular Docking which includes 1D and 3D rotations, have non-Euclidean search spaces. Does this influence or bias the results and search behavior of global optimizers (e.g. GA, PSO, DE, ES, MA...)? Are there suggestions on which search strategies or design decisions are more effective when solving this kind of problems?
.
I had encountered a somewhat similar problem (although at a much much more modest scale !) when faced with the apparently obvious problem of averaging rotations ... so as to implement some kind of k-means clustering on rotations ; i never completed this study but i had dug a few references (attached below)
.
.
As long as your space is a metric space, you can define differences. Note that for optimization purposes, your space only need to be partially ordered and that's it. The reason for partial order prerequisite is that being better need to be defined, and it is only defined if your space can be ordered. Also, if your space is not metric (differences are not measurable), then optimization is not defined as being better is not defined.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics