<div dir="ltr"><p class="" style="text-align:center"><span style="font-size:18pt;font-family:'Times New Roman',serif">SEMINARIO
DI FINANZA QUANTITATIVA</span></p>

<p class=""><span style="font-size:16pt;font-family:'Times New Roman',serif"> </span></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><span style="font-size:16pt;font-family:'Times New Roman',serif">Venerdì 11 marzo 2016</span><span style="font-size:16pt;font-family:'Times New Roman',serif"></span></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><span style="font-size:16pt;font-family:'Times New Roman',serif">ore 10:30</span></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><span style="font-size:16pt;font-family:'Times New Roman',serif"> </span></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><u><span style="font-size:16pt;font-family:'Times New Roman',serif">Scuola Normale Superiore</span></u></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><span style="font-size:16pt;font-family:'Times New Roman',serif">Pisa</span></p>

<p class="MsoNormal" align="center" style="margin-left:9pt;text-align:center;line-height:18pt"><span style="font-size:16pt;font-family:'Times New Roman',serif;background-image:initial;background-repeat:initial">Aula Bianchi Scienze</span></p>

<p class="MsoNormal" align="center" style="text-align:center"><b><u><span style="font-size:22pt;font-family:'Times New Roman',serif"><span style="text-decoration:none"> </span></span></u></b></p>

<p class="MsoNormal" align="center" style="text-align:center;background-image:initial;background-repeat:initial"><b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif">Christian Brownlees</span></b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif"></span></p>

<p class="MsoNormal" align="center" style="text-align:center;background-image:initial;background-repeat:initial"><i><span style="font-family:'Times New Roman',serif">(Universitat Pompeu
Fabra, Barcelona, Spain)</span></i></p>

<pre style="text-align:center"><b><i><span style="font-size:12pt;font-family:'Times New Roman',serif;background-image:initial;background-repeat:initial"> </span></i></b></pre>

<p class="MsoNormal" align="center" style="text-align:center"><span style="font-size:16pt;font-family:'Times New Roman',serif">terrà un
seminario dal titolo:</span></p>

<p class="MsoNormal" align="center" style="text-align:center"><span style="font-size:16pt;font-family:'Times New Roman',serif"> </span></p>

<p class="MsoNormal" align="center" style="text-align:center;background-image:initial;background-repeat:initial"><b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif;color:black">“</span></b><b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif;background-image:initial;background-repeat:initial">Community Detection in Partial Correlation Networks</span></b><b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif;color:black">”</span></b><b><span lang="EN-US" style="font-size:22pt;font-family:'Times New Roman',serif"></span></b></p>

<p class="MsoNormal" style="text-align:justify"><b><span lang="EN-US" style="font-family:'Times New Roman',serif"> </span></b></p>

<p class="MsoNormal" style="text-align:justify"><b><span lang="EN-US" style="font-family:'Times New Roman',serif">Abstract:</span></b></p>

<p class="MsoNormal" style="text-align:justify"><span lang="EN-US" style="font-family:'Times New Roman',serif;background-image:initial;background-repeat:initial">In this work we propose a community detection
algorithm for partial correlation networks. We assume that the variables
in the network are partitioned into communities. The presence of nonzero
partial correlation between two variables is determined by a Bernoulli trial
whose probability depends on whether the variables belong to the same
community or not. The community partition is assumed to be unobserved and
the goal is to recover it from a sample of observations. To tackle this
problem we introduce a community detection algorithm called
Blockbuster. The algorithm detects communities by applying k-means
clustering to the eigenvectors corresponding to the largest eigenvalues of the
sample covariance matrix. We study the properties of the procedure and
show that Blockbuster consistently detects communities when the network
dimension and the sample size are large. The methodology is used to study
real activity clustering in the United States</span><span lang="EN-US" style="font-size:9.5pt;font-family:Arial,sans-serif">.</span><span lang="EN-US" style="font-family:'Times New Roman',serif"></span></p>

<p class="MsoNormal" style="text-align:justify;background-image:initial;background-repeat:initial"><span lang="EN-US" style="font-family:'Times New Roman',serif"> </span></p>

<p class="MsoNormal" style="text-align:justify"><span style="font-size:16pt;font-family:'Times New Roman',serif">Tutti gli interessati sono invitati a
partecipare.</span></p>

<p class="MsoNormal" style="text-align:justify"><span style="font-size:16pt;font-family:'Times New Roman',serif"> </span></p>

<p class="MsoNormal" style="text-align:justify"><span style="font-size:16pt;font-family:'Times New Roman',serif">Classe di Scienze Matematiche e Naturali</span></p>

<p class="MsoNormal" style="text-align:justify"><span style="font-size:16pt;font-family:'Times New Roman',serif"> </span></p><div><div class="gmail_signature"><div dir="ltr"><div>Valeria Giuliani</div><div>Scuola Normale Superiore</div><div>Servizio alla Didattica e Allievi</div><div>tel. 050 509260</div><div>Piazza dei Cavalieri, 7</div><div>56126 Pisa</div><div>E-mail: <a href="mailto:valeria.giuliani@sns.it" target="_blank">valeria.giuliani@sns.it</a></div><div>E-mail: <a href="mailto:classi@sns.it" target="_blank">classi@sns.it</a></div></div></div></div>
</div>