<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.0.5" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments for Greg's Blog</title>
	<link>http://greggurevich.com</link>
	<description>To be in the Tech industry, you need to have a blog, here is mine. I promise to make it as boring as the rest of them, so you might as well leave now.                                        I am even using a cookie cutter Wordpress theme, so you know this won't be good.</description>
	<pubDate>Fri, 21 Nov 2008 13:15:28 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.0.5</generator>

	<item>
		<title>Comment on Better than rules, how Bayesian theory is being applied on the Net by Sam Danielson</title>
		<link>http://greggurevich.com/2007/04/19/the-rise-of-statistics/#comment-250</link>
		<pubDate>Fri, 20 Apr 2007 02:02:12 +0000</pubDate>
		<guid>http://greggurevich.com/2007/04/19/the-rise-of-statistics/#comment-250</guid>
					<description>I think there are bigger applications than spam and automatic classification. These Bayesian inferencing systems could be used to filter job applications, help you decide where to eat in a strange city, or control a helper robot. The sky is the limit. Unfortunately no one knows how well an application will work until it is run on real data and the data gathering is expensive. Look at how much time and effort PG put into gathering his test data and then tweaking the algorithm, all for what amounts to a science-project scale system. It's interesting work and I see it as more of a Thomas Edison problem than an Einstein one. One has to get their hands dirty.</description>
		<content:encoded><![CDATA[<p>I think there are bigger applications than spam and automatic classification. These Bayesian inferencing systems could be used to filter job applications, help you decide where to eat in a strange city, or control a helper robot. The sky is the limit. Unfortunately no one knows how well an application will work until it is run on real data and the data gathering is expensive. Look at how much time and effort PG put into gathering his test data and then tweaking the algorithm, all for what amounts to a science-project scale system. It&#8217;s interesting work and I see it as more of a Thomas Edison problem than an Einstein one. One has to get their hands dirty.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>Comment on Working Late by Administrator</title>
		<link>http://greggurevich.com/2005/11/15/working-late/#comment-249</link>
		<pubDate>Thu, 19 Apr 2007 17:39:49 +0000</pubDate>
		<guid>http://greggurevich.com/2005/11/15/working-late/#comment-249</guid>
					<description>I always hated posting code</description>
		<content:encoded><![CDATA[<p>I always hated posting code
</p>
]]></content:encoded>
				</item>
	<item>
		<title>Comment on Working Late by Robert sandie</title>
		<link>http://greggurevich.com/2005/11/15/working-late/#comment-3</link>
		<pubDate>Fri, 18 Nov 2005 03:28:11 +0000</pubDate>
		<guid>http://greggurevich.com/2005/11/15/working-late/#comment-3</guid>
					<description>You need to post the FFMPEG source that you were working on when you wrote this.</description>
		<content:encoded><![CDATA[<p>You need to post the FFMPEG source that you were working on when you wrote this.
</p>
]]></content:encoded>
				</item>
</channel>
</rss>
