image
imagewidth (px) 1.24k
1.76k
| markdown
stringlengths 1
996
|
|---|---|
<image>
|
|
<image>
LPG produces additional intermediate solutions, that for clarity are not shown in these plots.
It is not surprising that very often plan quality for LPG-speed is poor with respect to plan quality for LPG-quality, and that the CPU-time required by LPG-quality is much higher that the CPU-time required by LPG-speed. Things become less clear if we compare plan quality for LPG -speed (or CPU-time for LPG -quality) and plan quality for the speed version of the SuperPlanner (or CPU-time for the quality version of the SuperPlanner).
Given the any time nature of LPG, obviously there is a tradeoff between plan quality and speed. The more CPU-time the planner is allowed to run, the better the last solution generated. If we want to study this tradeoff experimentally, we need to consider not only the first and last solutions that LPG found in the competition tests, but also the interme- |
|
<image>
Appendix B: LPG and the SuperPlanner in the Time variant of the ompetition domains
|
|
<image>
|