We demonstrate that it is possible to achieve accurate localization and tracking of a target in a randomly placed wireless sensor network composed of inexpensive components of limited accuracy. The crucial enabler for this is a reasonably accurate local coordinate system aligned with the global coordinates. We present an algorithm for creating such a coordinate system without the use of global control, globally accessible beacon signals, or accurate estimates of inter-sensor distances. The coordinate system is robust and automatically adapts to the failure or addition of sensors. Extensive theoretical analysis and simulation results are presented. Two key theoretical results are: there is a critical minimum average neighborhood size of 15 for good accuracy and there is a fundamental limit on the resolution of any coordinate system determined strictly from local communication. Our simulation results show that we can achieve position accuracy to within 20% of the radio range even when there is variation of up to 10% in the signal strength of the radios. The algorithm improves with finer quantizations of inter-sensor distance estimates: with 6 levels of quantization position errors better than 10% are achieved. Finally we show how the algorithm gracefully generalizes to target tracking tasks.

/pdf/organizing-a-global-coordinate-system-from-local-information-2c479pqenf.pdf

Organizing a global coordinate system from local information on an ad hoc sensor network

Congressional roll call voting has been highly structured for most of U.S. history. The structure is revealed by a dynamic, spatial analysis of the entire roll call voting record from 1789 to 1985. The space is characterized by a predominant major dimension with, at times, a significant, but less important second dimension. In the modern era, spatial positions are very stable. This stability is such that, under certain conditions, short run forecasting of roll call votes is possible. Since the end of World War II, changes in congressional voting patterns have occurred almost entirely through the process of replacement of retiring or defeated legislators with new members. Politically, selection is far more important than adaptation.

/pdf/patterns-of-congressional-voting-1ibbdforzt.pdf

Patterns of congressional voting

Information retrieval (IR) researchers commonly use three tests of statistical significance: the Student's paired t-test, the Wilcoxon signed rank test, and the sign test. Other researchers have previously proposed using both the bootstrap and Fisher's randomization (permutation) test as non-parametric significance tests for IR but these tests have seen little use. For each of these five tests, we took the ad-hoc retrieval runs submitted to TRECs 3 and 5-8, and for each pair of runs, we measured the statistical significance of the difference in their mean average precision. We discovered that there is little practical difference between the randomization, bootstrap, and t tests. Both the Wilcoxon and sign test have a poor ability to detect significance and have the potential to lead to false detections of significance. The Wilcoxon and sign tests are simplified variants of the randomization test and their use should be discontinued for measuring the significance of a difference between means.

/pdf/a-comparison-of-statistical-significance-tests-for-7pu4t52tbf.pdf

A comparison of statistical significance tests for information retrieval evaluation

The Availability and Cost of Healthier Food Alternatives

Summary Background Listeriosis, caused by Listeria monocytogenes , is an important foodborne disease that can be difficult to control and commonly results in severe clinical outcomes. We aimed to provide the first estimates of global numbers of illnesses, deaths, and disability-adjusted life-years (DALYs) due to listeriosis, by synthesising information and knowledge through a systematic review. Methods We retrieved data on listeriosis through a systematic review of peer-reviewed and grey literature (published in 1990–2012). We excluded incidence data from before 1990 from the analysis. We reviewed national surveillance data where available. We did a multilevel meta-analysis to impute missing country-specific listeriosis incidence rates. We used a meta-regression to calculate the proportions of health states, and a Monte Carlo simulation to generate DALYs by WHO subregion. Findings We screened 11 722 references and identified 87 eligible studies containing listeriosis data for inclusion in the meta-analyses. We estimated that, in 2010, listeriosis resulted in 23 150 illnesses (95% credible interval 6061–91 247), 5463 deaths (1401–21 497), and 172 823 DALYs (44 079–676 465). The proportion of perinatal cases was 20·7% (SD 1·7). Interpretation Our quantification of the global burden of listeriosis will enable international prioritisation exercises. The number of DALYs due to listeriosis was lower than those due to congenital toxoplasmosis but accords with those due to echinococcosis. Urgent efforts are needed to fill the missing data in developing countries. We were unable to identify incidence data for the AFRO, EMRO, and SEARO WHO regions. Funding WHO Foodborne Diseases Epidemiology Reference Group and the Universite catholique de Louvain.

/pdf/the-global-burden-of-listeriosis-a-systematic-review-and-3830clvucx.pdf

R. L. Scheaffer

Papers

Mathematical Statistics with Applications