<oai_dc:dc xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
<dc:title>On orthogonal matrices maximizing the 1-norm</dc:title>
<dc:creator>Teodor Banica</dc:creator><dc:creator>Benoit Collins</dc:creator><dc:creator>Jean-Marc Schlenker</dc:creator>
<dc:subject>47A30</dc:subject><dc:subject>05B20</dc:subject><dc:subject>orthogonal group</dc:subject><dc:subject>Hadamard matrix</dc:subject>
<dc:description>For $U \in O(N)$ we have $\|U\|_1 \leq N\sqrt{N}$, with equality if and only if $U = H/\sqrt{N}$, with $H$ Hadamard matrix. Motivated by this remark, we discuss in this paper the algebraic and analytic aspects of the computation of the maximum of the 1-norm on $O(N)$. The main problem is to compute the $k$-th moment of the 1-norm on $O(N)$, with $k \to \infty$, and we discuss here the general strategy for approaching this problem, with some explicit computations at $k = 1$, $2$ and $N \to \infty$.</dc:description>
<dc:publisher>Indiana University Mathematics Journal</dc:publisher>
<dc:date>2010</dc:date>
<dc:type>text</dc:type>
<dc:format>pdf</dc:format>
<dc:identifier>10.1512/iumj.2010.59.3926</dc:identifier>
<dc:source>10.1512/iumj.2010.59.3926</dc:source>
<dc:language>en</dc:language>
<dc:relation>Indiana Univ. Math. J. 59 (2010) 839 - 856</dc:relation>
<dc:coverage>state-of-the-art mathematics</dc:coverage>
<dc:rights>http://iumj.org/access/</dc:rights>
</oai_dc:dc>