robocode OLAP experiment

[Pavel Šavara]

Hi All,

I have plan to ammend robocode/roborumble client, so it will produce
stream of data about every tick of game
(positions/energy/score/bullets/etc ..). Actually I will use latest
robocode 1.7.x and use record-battle functionality (I'm not worried
about 100% correctness of the data).

I will stream these data into some OLAP data cube. When I have enough
data I will run statistical analysis on it. I hope to find something
interesting, for example that there is better chance to hit robot in
some specific place on battlefield, or in specific situation. I'm data
mining newbie, I'm not sure yet what it will bring.

My current plan is
- build local roborumble server (by darcanuck)
- use just top 20 participants of general 1-on-1
- create OLAP database structures
- amend the robocode client to upload the data
- run it for while to gather the data
- help appreciated: to run the new client on your HW
- help appreciated: help me choose and design OLAP

Step 2
- build aggregation rules
- data processing: triggers ?, views ?, stored procedures ?
- types of aggregations / correlations ?
- help appreciated: data mining knowledge
- help appreciated: interesting robocode patterns to be explored


Anyone interested, please shout, I would appreciate help.
Thanks
Pavel Savara


------------------------------------

Yahoo! Groups Links

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/Robocode/

<*> Your email settings:
Individual Email | Traditional

<*> To change settings online go to:
http://groups.yahoo.com/group/Robocode/join
(Yahoo! ID required)

<*> To change settings via email:
Robocode-***@yahoogroups.com
Robocode-***@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
Robocode-***@yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/