The Battle for Wesnoth Forums

Hi Quitch,

Thanks for looking at this. I'd like to help out, but unfortunately I'm on a Mac and it sounds like your on Windows. Can anybody help out with the best way to apply a patch on Windows?

SD

Instructions for how to apply the patch can be found here: http://wiki.wesnoth.org/Machine_Learnin ... _the_patch.

As it says at the above link, you get the patch from https://gna.org/patch/?3479. Take the latest version, which is currently ML Recruiter 0.4.

I may need help with Windows and Linux since I only have access to a Mac, but I'd be happy to debug any issues.

With regards to the mail, could you add me to the dev mailing list?

Wow, we're going to have an unbeatable AI in a year's time if this goes on... I was thinking that, regarding the seemingly growing interest in experimental AI improvements, there could be a branch for this kind of stuff? I would be more eager to test patches myself if they didn't pollute my main wo...

BTree, Thanks for your comments about the ML Recruiter. I'm glad to hear that another machine learning person thinks I'm on the right track. I've just released the ML Recruiter 0.4 patch, which you can find at https://gna.org/patch/?3479. With this version, I am migrating the project to join in with...

For anyone who wants to learn more about ML Recruiter, I've just updated the Wiki at http://wiki.wesnoth.org/Machine_Learning_Recruiter . New items are faction vs. faction statistics for ML Recruiter 0.3, faction-by-faction unit recruitment percentages, a further breakdown of unit recruitment percen...

Those results sound very exciting. I like your improvements. Thanks! As you can see, I implemented your idea for a new metric and I think it helped. What would it take to make it work with those other maps? I'm pretty sure I know the issues: Thousand Stings and Caves of the Basilisk: These have pet...

ML Recruiter 0.3 is out and posted at https://gna.org/patch/index.php?3479 . The Machine Learning Recruiter is a brand new recruiting algorithm which achieves dramatically better performance than the current RCA recruiter through the use of the latest neural net technology. The ML Recruiter automati...

Just a comment that, like mattsc, I've been working on the AI too, focusing on the recruiting algorithm. You can read about it here: http://wiki.wesnoth.org/Machine_Learning_Recruiter . I'm taking a machine learning approach to the problem, having a neural net automatically learn its recruiting stra...

I'm writing a workaround for this bug for the default gold issue into the new version of ai_test2.py which is included with my ML Recruiter patch, http://wiki.wesnoth.org/Machine_Learning_Recruiter . The workaround will add --parm 1:gold:100 --parm 2:gold:100 to the command-line when the map is not ...

I'm also encountering the same problem with the experience modifier. I'd like to train and test the ML Recruiter (http://wiki.wesnoth.org/Machine_Learning_Recruiter) using a 70% experience modifier (which allows units to advance with only 70% of the experience usually required) since that's the defa...

An easy improvement to AI recruiting would be banking, i.e. not recruiting when possible but not necessary to save gold and upkeep to have more of it later. ... it might be well doable with a gold metric (wasted upkeep). As you say, this idea would come into play on one of those campaign scenarios ...

No, unfortunately that doesn't work. So this code seems to be doing the right thing. Am I missing something? if defender.status.not_living == true then print(defender.id .. " is not living") else print(defender.id .. " is living") end

So I'm beginning to code up a variant of Sapient's suggested goodness metric, which I'm calling the gold yield metric. Here's how I see it: Basic Damage Metric: Target unit cost * (Damage inflicted/target max HP) (Sapient's suggestion). The concept is that you cost your opponent this much gold by de...

local status = H.get_child(defender.__cfg, "status") local cant_poison = status.poisoned or status.not_living Thanks, that all seems to be working. However, wouldn't it be easier to write the above code snippet as: local cant_poison = defender.status.poisoned or defender.status.not_living