Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userMap Projections Minimizing Distance Errors
95
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
It is useful to have mathematical criteria for evaluating errors in map projections. The Chebyshev criterion for minimizing rms (root mean square) local scale factor errors for conformal maps has been useful in developing conformal map projections of continents. Any local error criterion will be minimized ultimately by map projections with multiple interruptions, on which some pairs of points that are close on the globe are far apart on the map. Since it is as bad to have two points on the map at two times their proper separation as to have them at half their proper separation, it is the rms logarithmic distance, s, between random points in the mapped region that we will minimize. The best previously known projection of the entire sphere for distances is the Lambert equal-area azimuthal with an rms logarithmic distance error of s=0.343. For comparison, the Mercator has s=0.444, and the Mollweide has s=0.390. We present new projections: the Gott equal-area elliptical with perfect shapes on the central meridian, the Gott-Mugnolo equal-area elliptical and the Gott-Mugnolo azimuthal with rms logarithmic distance errors of s=0.365, s=0.348, and s=0.341 respectively, which improve on previous projections of their type. The Gott-Mugnolo azimuthal has the lowest distance errors of any map and is produced by a new technique using forces between pairs of points on a map which make them move so as to minimize s. The Gott equal-area elliptical projection produces a particularly attractive map of Mars, and the Gott-Mugnolo azimuthal projection produces an interesting map of the moon.
rate research
Read More
This paper introduces a new way to calculate distance-based statistics, particularly when the data are multivariate. The main idea is to pre-calculate the optimal projection directions given the variable dimension, and to project multidimensional variables onto these pre-specified projection directions; by subsequently utilizing the fast algorithm that is developed in Huo and Szekely [2016] for the univariate variables, the computational complexity can be improved from $O(m^2)$ to $O(n m cdot mbox{log}(m))$, where $n$ is the number of projection directions and $m$ is the sample size. When $n ll m/log(m)$, computational savings can be achieved. The key challenge is how to find the optimal pre-specified projection directions. This can be obtained by minimizing the worse-case difference between the true distance and the approximated distance, which can be formulated as a nonconvex optimization problem in a general setting. In this paper, we show that the exact solution of the nonconvex optimization problem can be derived in two special cases: the dimension of the data is equal to either $2$ or the number of projection directions. In the generic settings, we propose an algorithm to find some approximate solutions. Simulations confirm the advantage of our method, in comparison with the pure Monte Carlo approach, in which the directions are randomly selected rather than pre-calculated.
Cartograms are maps that rescale geographic regions (e.g., countries, districts) such that their areas are proportional to quantitative demographic data (e.g., population size, gross domestic product). Unlike conventional bar or pie charts, cartograms can represent correctly which regions share common borders, resulting in insightful visualizations that can be the basis for further spatial statistical analysis. Computer programs can assist data scientists in preparing cartograms, but developing an algorithm that can quickly transform every coordinate on the map (including points that are not exactly on a border) while generating recognizable images has remained a challenge. Methods that translate the cartographic deformations into physics-inspired equations of motion have become popular, but solving these equations with sufficient accuracy can still take several minutes on current hardware. Here we introduce a flow-based algorithm whose equations of motion are numerically easier to solve compared with previous methods. The equations allow straightforward parallelization so that the calculation takes only a few seconds even for complex and detailed input. Despite the speedup, the proposed algorithm still keeps the advantages of previous techniques: with comparable quantitative measures of shape distortion, it accurately scales all areas, correctly fits the regions together and generates a map projection for every point. We demonstrate the use of our algorithm with applications to the 2016 US election results, the gross domestic products of Indian states and Chinese provinces, and the spatial distribution of deaths in the London borough of Kensington and Chelsea between 2011 and 2014.
Deep neural networks (DNNs) are poorly calibrated when trained in conventional ways. To improve confidence calibration of DNNs, we propose a novel training method, distance-based learning from errors (DBLE). DBLE bases its confidence estimation on distances in the representation space. In DBLE, we first adapt prototypical learning to train classification models. It yields a representation space where the distance between a test sample and its ground truth class center can calibrate the models classification performance. At inference, however, these distances are not available due to the lack of ground truth labels. To circumvent this by inferring the distance for every test sample, we propose to train a confidence model jointly with the classification model. We integrate this into training by merely learning from mis-classified training samples, which we show to be highly beneficial for effective learning. On multiple datasets and DNN architectures, we demonstrate that DBLE outperforms alternative single-model confidence calibration approaches. DBLE also achieves comparable performance with computationally-expensive ensemble approaches with lower computational cost and lower number of parameters.
Convolutional neural networks for semantic segmentation suffer from low performance at object boundaries. In medical imaging, accurate representation of tissue surfaces and volumes is important for tracking of disease biomarkers such as tissue morphology and shape features. In this work, we propose a novel distance map derived loss penalty term for semantic segmentation. We propose to use distance maps, derived from ground truth masks, to create a penalty term, guiding the networks focus towards hard-to-segment boundary regions. We investigate the effects of this penalizing factor against cross-entropy, Dice, and focal loss, among others, evaluating performance on a 3D MRI bone segmentation task from the publicly available Osteoarthritis Initiative dataset. We observe a significant improvement in the quality of segmentation, with better shape preservation at bone boundaries and areas affected by partial volume. We ultimately aim to use our loss penalty term to improve the extraction of shape biomarkers and derive metrics to quantitatively evaluate the preservation of shape.
begin{abstract} In this paper we consider Time-Varying Block (TVB) codes, which generalize a number of previous synchronization error-correcting codes. We also consider various practical issues related to MAP decoding of these codes. Specifically, we give an expression for the expected distribution of drift between transmitter and receiver due to synchronization errors. We determine an appropriate choice for state space limits based on the drift probability distribution. In turn, we obtain an expression for the decoder complexity under given channel conditions in terms of the state space limits used. For a given state space, we also give a number of optimizations that reduce the algorithm complexity with no further loss of decoder performance. We also show how the MAP decoder can be used in the absence of known frame boundaries, and demonstrate that an appropriate choice of decoder parameters allows the decoder to approach the performance when frame boundaries are known, at the expense of some increase in complexity. Finally, we express some existing constructions as TVB codes, comparing performance with published results, and showing that improved performance is possible by taking advantage of the flexibility of TVB codes.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
