home

about us

articles

software

get involved

 

Articles

The Initiative for Interdisciplinary Research has a wide range of research activities within the social sciences. The articles and reports below outline some of the research currently being done. Any questions or comments should be relayed to Wojciech Gryc.

Self-Correlation Analysis of the Photometric Variability of T Tauri Stars

T Tauri stars are variable stars that are in an early phase of evolution, in which accretion and contraction to the main sequence are still taking place. Their photometric variability is complex; it takes place on a variety of timescales, due to a variety of physical processes. Periodic variability occurs due to rotation and the presence of cool or hot spots on the star. It may also occur due to periodic obscuration of the star by inhomogeneities in the still present accretion disk. But the periodicity may be masked by other forms of variability, or by time variation in the cool or hot spots, or the obscuring inhomogeneities. For other types of variable stars, self-correlation has proven to be a useful adjunct to Fourier analysis for studying semiregular variability; it determines the cycle-to-cycle behavior of the star, averaged over all the data. We have therefore used it to investigate the photometric variability of about 30 T Tauri stars using existing data. It has provided useful information about periods and their coherence, about the amplitude of the periodic variation, or its upper limit, and about the “profile” of the amplitude-timescale behavior. In most cases, it has confirmed periods previously determined by Fourier analysis, but in some cases it has suggested that the previously determined period is spurious.

Published in the Publications of the Astronomical Society of the Pacific, October 2006 (Volume 118).

Neural Network Predictions of Stock Price Fluctuations

This research paper is the result of a three-month-long independent study focusing on the ability of feedforward and recurrent neural networks in predicting stock price fluctuations of companies within the "Consumer Discretionary" category of the Global Industry Classification Standard (GICS). The paper focuses on predictions for Circuit City Stores Inc. (CC) for 1-, 5-, and 20-business day periods through the use of input patterns based on fundamental and technical indicators from the stock market and economy. The project is unique in its focus on companies selling to consumers. The results of the project are promising, with statistically significant results appearing for 1-day single-layer feedforward and Elman networks, as well as networks based on fundamental inputs and predicting for 20-day periods. A PDF version of a similar PowerPoint presentation is also available.

Intergroup Conflict: A social network approach

Network-based analysis of intergroup conflict has shown promising results with regards to understanding how conflicts develop within society. This paper gives a general overview of how one can apply structural and dynamic network approaches to shed light on conflict, and establishes a qualitative description of how mathematical concepts may be used to describe such phenomena. Two conflicts are studied within the context of the social network approach: the al Qaeda terrorist network, as well as the genocide in Rwanda. A social network approach to intergroup conflict appears to be a useful method of analyzing conflicts, and helps shed light on their prevention.

Al Qaeda: Corporate hierarchy or biological entity?

A great deal of study has been devoted to network-centric organizations in society, and terrorism is no exception. This article explores Al Qaeda'a organizational structure and tries to deduce its most likely shape based on three choices: hierarchy, network, and chaotic. While there are supporting arguments for all three structures, there is a promising research with regards to viewing Al Qaeda as a dynamic and chaotic organization that has little, if any, formal structure. This may be the most exciting development presented in this paper, and begs to ask the question, "How does one target an entity with independent nodes and no formal structure?"