<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:georss="http://www.georss.org/georss" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:media="http://search.yahoo.com/mrss/"
		>
<channel>
	<title>Comments on: LDA Clustering for Gene Pathway Inference</title>
	<atom:link href="http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/feed/" rel="self" type="application/rss+xml" />
	<link>http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/</link>
	<description>Natural Language Processing and Text Analytics</description>
	<lastBuildDate>Sat, 04 Feb 2012 20:56:48 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.com/</generator>
	<item>
		<title>By: lingpipe</title>
		<link>http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/#comment-6480</link>
		<dc:creator><![CDATA[lingpipe]]></dc:creator>
		<pubDate>Mon, 01 Mar 2010 21:16:28 +0000</pubDate>
		<guid isPermaLink="false">http://lingpipe-blog.com/?p=3760#comment-6480</guid>
		<description><![CDATA[My bad for not citing that paper to begin with.  I even saw you give the talk on it at NYU!  I think my problem was that I didn&#039;t understand multivariate normal priors and their relation to multinomials until after reading Gelman and Hill&#039;s book and mulling over Gelman&#039;s comment about why he didn&#039;t like Dirichlet priors.  

What&#039;d be nice is to combine this idea with predictors for the pools in a general multilevel model. 

P.S. The WordPress spam detector doesn&#039;t like links, which is why I had to manually approve this one.]]></description>
		<content:encoded><![CDATA[<p>My bad for not citing that paper to begin with.  I even saw you give the talk on it at NYU!  I think my problem was that I didn&#8217;t understand multivariate normal priors and their relation to multinomials until after reading Gelman and Hill&#8217;s book and mulling over Gelman&#8217;s comment about why he didn&#8217;t like Dirichlet priors.  </p>
<p>What&#8217;d be nice is to combine this idea with predictors for the pools in a general multilevel model. </p>
<p>P.S. The WordPress spam detector doesn&#8217;t like links, which is why I had to manually approve this one.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: david blei</title>
		<link>http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/#comment-6475</link>
		<dc:creator><![CDATA[david blei]]></dc:creator>
		<pubDate>Mon, 01 Mar 2010 01:36:32 +0000</pubDate>
		<guid isPermaLink="false">http://lingpipe-blog.com/?p=3760#comment-6475</guid>
		<description><![CDATA[Hi Bob

Your gripe with the Dirichlet was exactly our motivation for the &quot;correlated topic model.&quot;  It uses a multivariate gaussian on the logit scale, which is called the logistic normal.  (See the excellent work of Aitchison.)  Inference is complicated because we no longer enjoy conjugacy between the topic proportions and topic indicators.  But the model is more expressive.

If interested, see the paper here:
  http://www.cs.princeton.edu/~blei/papers/BleiLafferty2007.pdf

There is C code here
  http://www.cs.princeton.edu/~blei/ctm-c/index.html

(Forgive the self promotion...)

Best,
Dave]]></description>
		<content:encoded><![CDATA[<p>Hi Bob</p>
<p>Your gripe with the Dirichlet was exactly our motivation for the &#8220;correlated topic model.&#8221;  It uses a multivariate gaussian on the logit scale, which is called the logistic normal.  (See the excellent work of Aitchison.)  Inference is complicated because we no longer enjoy conjugacy between the topic proportions and topic indicators.  But the model is more expressive.</p>
<p>If interested, see the paper here:<br />
  <a href="http://www.cs.princeton.edu/~blei/papers/BleiLafferty2007.pdf" rel="nofollow">http://www.cs.princeton.edu/~blei/papers/BleiLafferty2007.pdf</a></p>
<p>There is C code here<br />
  <a href="http://www.cs.princeton.edu/~blei/ctm-c/index.html" rel="nofollow">http://www.cs.princeton.edu/~blei/ctm-c/index.html</a></p>
<p>(Forgive the self promotion&#8230;)</p>
<p>Best,<br />
Dave</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: tmasada</title>
		<link>http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/#comment-6462</link>
		<dc:creator><![CDATA[tmasada]]></dc:creator>
		<pubDate>Fri, 26 Feb 2010 14:23:50 +0000</pubDate>
		<guid isPermaLink="false">http://lingpipe-blog.com/?p=3760#comment-6462</guid>
		<description><![CDATA[Simon Rogers, Mark Girolami, Colin Campbell and Rainer Breitling. The latent process decomposition of cDNA microarray data sets. IEEE/ACM Transactions on Computational Biology and Bioinformatics. Vol. 2, No. 2. pp. 143-156. 2005.

Yiming Ying, Peng Li and Colin Campbell. A marginalized variational Bayesian approach to the analysis of array data. BMC Proceedings 2008, 2(Suppl 4):S7.
... This paper proposes a collapsed variational Bayesian inference for latent process decomposition.]]></description>
		<content:encoded><![CDATA[<p>Simon Rogers, Mark Girolami, Colin Campbell and Rainer Breitling. The latent process decomposition of cDNA microarray data sets. IEEE/ACM Transactions on Computational Biology and Bioinformatics. Vol. 2, No. 2. pp. 143-156. 2005.</p>
<p>Yiming Ying, Peng Li and Colin Campbell. A marginalized variational Bayesian approach to the analysis of array data. BMC Proceedings 2008, 2(Suppl 4):S7.<br />
&#8230; This paper proposes a collapsed variational Bayesian inference for latent process decomposition.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: sth4nth</title>
		<link>http://lingpipe-blog.com/2010/02/25/lda-clustering-for-gene-pathway-inference/#comment-6460</link>
		<dc:creator><![CDATA[sth4nth]]></dc:creator>
		<pubDate>Fri, 26 Feb 2010 09:10:49 +0000</pubDate>
		<guid isPermaLink="false">http://lingpipe-blog.com/?p=3760#comment-6460</guid>
		<description><![CDATA[D. Blei, T. Griffiths, and M. Jordan.   The nested Chinese restaurant process and Bayesian nonparametric inference of topic hierarchies.   Journal of the ACM 2010

Y. W. Teh, M. I. Jordan, M. J. Beal and D. M. Blei.  Hierarchical Dirichlet processes.  Journal of the American Statistical Association, 101, 1566-1581, 2006.]]></description>
		<content:encoded><![CDATA[<p>D. Blei, T. Griffiths, and M. Jordan.   The nested Chinese restaurant process and Bayesian nonparametric inference of topic hierarchies.   Journal of the ACM 2010</p>
<p>Y. W. Teh, M. I. Jordan, M. J. Beal and D. M. Blei.  Hierarchical Dirichlet processes.  Journal of the American Statistical Association, 101, 1566-1581, 2006.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

