The event "Conference and Summer School: Algebraic Statistics in the Alleghenies" will be held at the Pennsylvania State University, University Park, PA, from June 9 to 15, 2012 (URL: www.math.psu.edu/morton/aspsu2012/). Algebraic statistics exploits algebraic geometry and related fields to solve problems in statistics and its applications. Methods from algebraic statistics have been successfully applied to address many problems including construction of Markov bases, theoretical study of phylogenetic mixture models, ecological inference, identifiability problems for graphical models, Bayesian integrals and singular learning theory, social networks, and coalescent theory. In addition to algebraic statistics' successes in solving statistical problems, its research objectives have driven theoretical developments in algebra.
Traditionally, applied mathematics has focused on branches of mathematics including differential equations and analysis. Instead, algebraic statistics advocates algebraic geometry, a well-developed and ancient field of mathematics, as a tool for solving problems in statistics and its applications. Many statistical models have the structure of algebraic varieties. This observation catalyzed rapid growth in this area over the past fifteen years. Over this time it has become clear that algebraic structures are ubiquitous in statistics. Hence advanced tools from algebra profitably address statistical questions. The purpose of this grant is to support a seven day conference and summer school Algebraic Statistics in the Alleghenies at the Pennsylvania State University, June 9-15, 2012. More than 100 participants are anticipated, which would make this the largest meeting yet on Algebraic Statistics that had been held in the US or abroad.
Pennsylvania State University hosted a large algebraic statistics meeting June 8 to June 15, 2012. Statistics is the science concerned with collecting, organizing, visualizing, analyzing and interoperating data, and drawing inferences about underlying processes and populations of interests. Algebra is the branch of mathematics concerned with discrete structures and finite computations. Algebraic statistics exploits the often discrete nature of data to turn many statistical problems into algebra problems, where advanced mathematical machinery can then be applied. Methods from algebraic statistics have been successfully applied to address many problems including construction of Markov bases, theoretical study of phylogenetic mixture models, ecological inference, identifiability problems for graphical models, Bayesian integrals and singular learning theory, social networks, and coalescent theory. In addition to algebraic statistics' successes in solving statistical problems, its research objectives have driven theoretical developments in algebra. The organizing committee of the conference (the PIs) welcomed contributions from interested participants, in form of contributed talks and posters, on the topics of methods and applications of algebraic statistics broadly defined, including but not limited to the above topics. The workshop also had tutorial lectures Saturday and Sunday, with the main conference Monday through Friday including both the longer plenary lectures below and short talks contributed by invitation. There were 4 tutorial lectures, 31 talks, and 16 posters at the poster session, and a total of approximately one hundred participants. Outcomes of the conference include many new collaborations, some aspect of which can be summarized as follows: - Two issues will published in the special volume of Journal of Algebraic Statistics (first issue came out in October 2012; second issue will appear in June 2013) - Conference was attended by international researchers including from Japan and various European countries. This included a visitor working on climate research who got a special grant to attend this conference to start new collaborations - Conference inspired organization of a European algebraic statistics group meeting which took place in September 2012 in Austria (organizers: Caroline Uhler and Ruriko Yoshida, participants of ASPSU2012).
