The PI is co-organizing a conference on long-range dependence, self-similarity, and heavy tails, which will take place in Research Triangle Park, North Carolina in April 2012. More details can be found on the conference web-site http://lrd2012.web.unc.edu. The three themes of the conference are studied and arise across a wide range of disciplines and applications, including probability, statistics, geophysics, telecommunications, engineering, economics, finance and insurance. Many of these fields were literally transformed by the use of models involving long-range dependence and heavy tails. For example, with a dramatic shift from circuit to packet switching in modern communication networks, these models took over the dominant place from the models based on Poisson assumptions and exponential distributions. New exciting areas of application such as social networks are posing new challenges where power laws, scaling limits, and self-similarity phenomena emerge in a new light.
The conference will feature talks from world renowned experts as well as poster presentations by students and junior researchers. It is expected to have an important and long lasting scientific, educational, and socioeconomic impact. The conference will foster the exchange of ideas between probabilists, theoretical and applied statisticians, other scientists, as well as students and practitioners. This will likely lead to formulating new problems relevant to important applications, thereby stimulating the development of new fundamental theoretical research. Many of these problems are expected to have a broad educational, scientific, and general impact on society. For example, in connection to understanding risk, gauging uncertainty in complex setting, or taking advantage of emerging network structures. Most importantly, the event would introduce young researchers and graduate students to important areas of research.
This is the final report on the conference "Long-Range Dependence, Self-Similarity and Heavy Tails" partially supported by the National Science Foundation under conference grant 1208965. The conference was held from April 19 – 21, 2012 at the Radisson Hotel in Research Triangle Park, NC. The scientific committee consisted of Richard Davis (Columbia University), Claudia Klüppelberg (Technical University Munich), Gennady Samorodnitsky (Cornell University) and Walter Willinger (AT&T Labs Research). Organizational issues were handled by Stilian Stoev (University of Michigan) and Vladas Pipiras (University of North Carolina). The conference featured 28 invited speakers and a poster session with 19 presenters. Each invited speaker gave a 45-minute presentation on a topic related to the main themes of the conference. The poster session took place in the evening of the first day of the conference. Most presenters in the poster session were graduate students and young researchers. The total number of registered participants (including speakers and poster presenters) was 73. A number of local faculty and graduate students have dropped by the conference without registering. A website for the conference was created and is still maintained at http://lrd2012.web.unc.edu/, containing the list of participants, speakers, poster presenters, sponsors, the program, talk slides and other information. Besides the National Science Foundation, partial financial support was also provided by the Army Research Office, Boston University, Cornell University, and the National Security Agency. Non-financial support was provided by Bernoulli Society and SAMSI. Part of the costs were covered by a registration fee. The National Science Foundation support was primarily used to offset the travel and lodging costs for junior participants. The conference brought together top experts from the fields of probability and statistics together with students and practitioners who study theoretical and applied problems related to long-range dependence, self-similarity and heavy tails. Our goal was to keep the interdisciplinary spirit of the conference. Many of the invited speakers are working at the interface, of Probability/Statistics, with other disciplines, such as Computer Science, Physics, Engineering, Environmental Sciences, Mathematical Finance, etc. A broad range of applied and theoretical topics was covered, including models for Internet traffic, financial applications, medical ultrasound, mathematical physics; statistical inference questions for long-range dependent and heavy-tailed data; theoretical aspects of fractional processes; analysis of extreme events. Emphasis was on expository lectures accessible to researchers with different backgrounds, or from different fields. The event thus provided an excellent venue for graduate students and young researchers to get a broad view of current challenges and promising directions of research in the field. The conference also fostered the exchange of ideas between probabilists, theoretical and applied statisticians, other scientists, as well as students and practitioners. This will likely lead to formulating new problems relevant to important applications, and thereby stimulate the development of new fundamental theoretical research. The conference has thus had a broad scientific, social and educational impact. From our informal conversations with participants after the conference, the event was considered a huge success.
