[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

Method/Function to Parse an XGID, api to XG?

Posted By: RobertFontaine
Date: Friday, 28 August 2015, at 11:52 p.m.

I have the schema that Xavier put up on his site for the XGID...

1. Has anyone done written a procedure to parse and XGID into its parts? (and the reverse) - Where can I find the code snippets? (I'm lazy).

I'm assembling a relational database for the first 3 rolls extensible to whatever. My head space is generally in creating busineess apps so I'm a lot more comfortable normalizing my data and driving it with sql and data mining it to start.

I'm prototyping my database.. I'm going to put a few web pages in front of it for presentation/manual data entry and will likely host the data with MySQL and put a web service in front of it for anyone who wanst to query the data directly or add more data.

What I would like to do once I have the data in place is....

I have a couple of xeon phi cards on my bench, and a couple of 8 core xeons... still need motherboard and ram. (effectively a 100 core linux box) This will give me a very nice little machine learning workstation to run 'R', numpy, et al on to do some investigation of the early play in search of Features.

Rather than using deep neural networks to learn the backgammon moves I'd like to use dnn to attempt to analyze the common features of the positions and play that we human folk can leverage these features in our OTB analysis.

2nd move characteristics for the reasonable moves have shaken to a pretty understandable set. 3rd moves still seem to be in the realm of Quiz Factor in many cases.

My math is worse than average so this is an opportunity for me to refresh my calc, linear alg, numerical methods AND work through Andrew NG's machine learning course.

I've been collecting hardware from the money my wife lets me keep for about 6 months with the plan to extend my data analysis skill. This seems like a good way to combine my useless interests. I'm a few months away from the workstation for analysis yet but I would like to establish my data and start thinking about how initial feature sets can be established for analysis.

I know there are a few scientific/mathematic folk on the board here and I am all ears for suggestions.

Messages In This Thread

 

Post Response

Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:

 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.