Exact Misclassification Probabilities for Plug-In Normal Quadratic Discriminant Functions: II. The Heterogeneous Case

We consider the problem of discriminating between two independent multivariate normal populations, N_p(μ₁, Σ₁) and N_p(μ₂, Σ₂), having distinct mean vectors μ₁ and μ₂ and distinct covariance matrices Σ₁ and Σ₂. The parameters μ₁, μ₂, Σ₁, and Σ₂ are unknown and are estimated by means of independent random training samples from each population. We derive a stochastic representation for the exact distribution of the “plug-in” quadratic discriminant function for classifying a new observation between the two populations. The stochastic representation involves only the classical standard normal, chi-square, and F distributions and is easily implemented for simulation purposes. Using Monte Carlo simulation of the stochastic representation we provide applications to the estimation of misclassification probabilities for the well-known iris data studied by Fisher (Ann. Eugen.7 (1936), 179–188); a data set on corporate financial ratios provided by Johnson and Wichern (Applied Multivariate Statistical Analysis, 4th ed., Prentice–Hall, Englewood Cliffs, NJ, 1998); and a data set analyzed by Reaven and Miller (Diabetologia16 (1979), 17–24) in a classification of diabetic status.     

An empirical study on the benefit of split loads with the pickup and delivery problem

Splitting loads such that the delivery of certain loads is completed in multiple trips rather than one trip has been shown to have benefit for both the classic Vehicle Routing Problem (VRP) and the Pickup and Delivery Problem (PDP). However, the magnitude of the benefit may be affected by various problem characteristics. In this paper, we characterize those real world environments in which split loads are most likely to be beneficial. Based on practitioner interest, we determine how the benefit is affected by the mean load size and variance, number of origins relative to the number of destinations, the percentage of origin–destination pairs with a load requiring service, and the clustering of origin and destination locations. We find that the magnitude of benefit is greatest for load sizes just over one half vehicle capacity as these loads can not be combined without splitting, while they are the easiest to combine on a vehicle with splitting; increases as the number of loads sharing an origin or destination increases because there are more potential load combinations to split at each stop; and increases as the average distance from an origin to a destination increases because splitting loads reduces the trips from origins to destinations.     

Unsolvable Problems About Small Cancellation and Word Hyperbolic Groups

We apply a construction of Rips to show that a number of algorithmicproblems concerning certain small cancellation groups and, inparticular, word hyperbolic groups, are recursively unsolvable.Given any integer k > 2, there is no algorithm to determinewhether or not any small cancellation group can be generatedby either two elements or more than k elements. There is a smallcancellation group E such that there is no algorithm to determinewhether or not any finitely generated subgroup of E is all ofE, or is finitely presented, or has a finitely generated secondintegral homology group.     

The structure of singly-periodic minimal surfaces

The research described in this paper was supported by research grant DE-FG02-86ER250125 of the Applied Mathematical Science subprogram of Office of Energy Research, U.S. Department of Energy, and National Science Foundation grants DMS-8611574 and DMS-8802858     

A technique for flow-enhanced magnetic resonance angiography of the lower extremities

A two-dimensional, flow-enhanced gradient echo pulse sequence for nuclear magnetic resonance angiography is described. It employs interleaved, presaturated slices to acquire data efficiently on imagers which favor interleaved acquisition over sequential acquisition for multislice imaging. It is useful on any imager when the effective TR is extended to enhance the sensitivity to slow flow. The technique was applied to the region from aortic bifurcation to the iliac bifurcations of three normal volunteers. The right and left common iliac arteries and veins, the separation of the external and internal iliac arteries, and secondary branches were clearly depicted.     

Scaling theory of the paranematic nematic critical point

A thermodymanic scaling analysis of the paranematic-nematic critical point is presented. Although presumed to be in the Ising universality class, general arguments suggest that the order parameter approaches the critical point in an asymmetrical manner in contrast to mean-field predictions.