793 / 1228
Inverse Optimisation and Planning
A Multi-Objective (MO) Problem
Bi-Objective Pareto Fronts
obtained for
22 prostate implants
. The variety shows that a
single objective optimization with constant importance factors does not give always a good
result.
In general a strong trade-off is observed.
*
*Lahanas, Milickovic, Baltas, Zamboglou: “Application of Multiobjective Evolutionary Algorithms for Dose Optimization Problems in
Brachytherapy”,
EMO 2001, LNCS 1993, 574-587, 2001.
Conformity
Homogeneity
PTV
Example of a Pareto Front:
there is
a
strong
trade-off
between the objectives/ objective
functions
f
1
and
f
2
.
The smaller the
f
1
value is that
we want the larger is the corresponding
f
2
value
.
The
“ideal point/plan” I
lies far away form the
front. There is a high dependence on the
selection of the
f
1
value.