OrcaM - A
NeuralTargeting robot by
PEZ
- V0.2 - Now only NeuralTargeting, now the source packaged with it is complete.
- V0.1 - First release
What's special about it?
Ummm, well,
NeuralTargeting is still special isn't it? =) It's open source with the hope that you will help make it better.
Great, I want to try it. Where can I download it?
https://www.robocoderepository.com/BotDetail.jsp?id=1388
How competetive is it?
Noone knows yet. It gives some top bots a fight though. Its closest relative,
Orca, gets beaten badly on the
EternalRumble.
How does it move?
RandomMovement
How does it fire?
It uses several
NeuralNetworks in a
VirtualGuns type of set up. Choosing the best network based on recent hit rate. It holds two networks in training all the time until it consders all networks are finished training.
It doesn't really. Though it's
RandomMovement uses the enemy fire power as a parameter.
How does the melee strategy differ from one-on-one strategy?
This is currently a
OneOnOne bot, though I think it should be very possible to make it work in
Melee (to the extent it can be said it works in
OneOnOne at least).
How does it select a target to attack/avoid in melee ?
If it sees it it will choose it.
What does it save between rounds and matches?
It remembers its neural nets between rounds. Nothing is saved between matches.
Where did you get the name?
Orca is it's closest relative and the M is for multi/mega/massive (take your pick).
Can I use your code?
Please do. It's included in the jar. And the main bot code is available on this wiki as well:
But you must make your bot open source as well if you use any major portion of the code. And you must share any improvements on this wiki.
What's next for your robot?
- Saving networks between rounds should, in theory, be a good thing.
- Lots of experimentations with different NN inputs and such.
- Subclassing NNCandidate so that different types of networks are used.
I bet it will have major problems with
DuelistNano, like all my bots. =)
What other robot(s) is it based on?
Orca and
Gouldingi
Comments