Modeling Urban Land Use Changes Using Support Vector Machines          下载免费PDF全文

Mileva Samardžić‐Petrović  Suzana Dragićević  Miloš Kovačević  Branislav Bajat 《Transactions in GIS》2016,20(5):718-734

Support Vector Machines (SVM) is a machine learning (ML) algorithm commonly applied to the classification of remotely sensing data and more recently for modeling land use changes. However, in most geospatial applications the current literature does not elaborate on specifications of the SVM method with respect to data sampling, attribute selection and optimal parameters choices. Therefore the main objective of this study is to present and investigate the SVM technique for modeling urban land use change. The SVM model building procedure is presented together with the detailed evaluation of the output results with respect to the choice of datasets, attributes and the change of SVM parameters. Geospatial datasets containing nine land use classes and spatial attributes for the Municipality of Zemun, Republic of Serbia were used for years 2001, 2003, 2007 and 2011. The Correlation‐based Feature Subset method, kappa coefficient, Area Under Receiver Operating Characteristic Curve (AUC) and kappa simulation were used to perform the model evaluation and compare the model outputs with the real land use datasets. The obtained results indicate that the SVM‐based models perform better when implementing balanced data sampling, reduced data sets to informative subsets of attributes and properly identify the optimal learning parameters.  相似文献   

A note on the thermal component of the equation of state in solids     

Vladan čelebonović 《Earth, Moon, and Planets》1991,54(2):145-149

A simple method for determining the thermal component of the EOS of solids under high pressure is proposed. Application to the interior of the Earth gives results in agreement with recent geophysical data.  相似文献   

Automatic grid azimuth by hour angle of the sun,a star or a planet using an electronic theodolite Kern E2     

Nikola Solarić 《Astrophysics and Space Science》1991,177(1-2):175-179

The paper deals with the automatic grid azimuth determination of an object on the earth using the electronic theodolite Kern E2 by observing the sun, a star, or a planet. The observation time and the readings of horizontal and vertical circles of the electronic theodolite enter automatically the electronic calculator Hewlett-Packard HP41CX. The calculator computes the grid azimuth of the terrestrial object and directs the observer what to do. Therefore, even a personnel without specific high education can make observations. For the solar observations it is not necessary to have an astronomical almanac, because the program computes the solar coordinates. The general input data, together with the measured ones, can be stored on a magnetic cassette and later on, if necessary, one can correct the general data and recalculate the azimuth. The method proved to be very practical in the field work.  相似文献   

Minor planet short periodic perturbations: The indirect part of the disturbing function     

Zoran Knežević 《Celestial Mechanics and Dynamical Astronomy》1993,55(4):387-404

Leverrier's development of the indirect part of the disturbing function has been extended to include terms up to degree 4 in eccentricity and inclination; the resulting series has been expressed with respect to a fixed plane, and in a computer readable form (a list of integers). Tests have been performed for the relative significance of the terms of degrees 2, 3 and 4, and estimates have been obtained for the accuracy of the short periodic perturbations of a minor planet, and of the corresponding mean orbital elements. It was found that: (i) even in extreme cases, the indirect part of the disturbing function gives rise to very small short periodic perturbations; (ii) bodies of very high eccentricity/inclination and those close to mean motion resonances are most significantly affected; (iii) indirect perturbations for minor planets can be computed up to the degree 2 terms only, without any significant loss of accuracy; and (iv) higher degree indirect perturbations appear to be important only for their contribution to the long periodic effects of higher order (with respect to the perturbing mass).  相似文献   

A model of neptune according to the Savić-Kašanin theory     

Vladan Čelebonović 《Earth, Moon, and Planets》1983,29(2):185-189

A model of the internal structure of Neptune has been calculated according to the Savi?-Ka?anin theory of behaviour of materials under high pressure.  相似文献   

High-pressure phase transitions — examples of classical predictability     

Vladan čelebonović 《Earth, Moon, and Planets》1992,58(3):203-213

We present 19 examples of materials whose high pressure phase transition points can be determined within a particular classical theory of dense matter. Theoretical predictions are compared with experimental results, and some possible causes of discrepancies are discussed.  相似文献   

A new orbit of binary 15 Monocerotis     

Z. Cvetković  I. Vince  S. Ninković 《New Astronomy》2010,15(3):302-306

  相似文献   

A visual basic spreadsheet macro for geochemical background analysis   总被引：3，自引：0，他引：3  

Nakić Z  Posavec K  Bacani A 《Ground water》2007,45(5):642-647

A Visual Basic macro entitled BACKGROUND calculates geochemical background values of chemical parameters and estimates threshold values separating background data from anomalies. The macro uses two statistical methods, the iterative 2-sigma technique and the calculated distribution function, and integrates these model-based objective methods into a widely accessible platform (i.e., MS Excel). The macro offers the possibility for automated processing of geochemical data and enables an automated generation of background range and threshold values for chemical parameters.  相似文献   

Formation of Coronal Shock Waves     

S. Lulić  B. Vršnak  T. Žic  I. W. Kienreich  N. Muhr  M. Temmer  A. M. Veronig 《Solar physics》2013,286(2):509-528

Magnetosonic wave formation driven by an expanding cylindrical piston is numerically simulated to obtain better physical insight into the initiation and evolution of large-scale coronal waves caused by coronal eruptions. Several very basic initial configurations are employed to analyze intrinsic characteristics of MHD wave formation that do not depend on specific properties of the environment. It turns out that these simple initial configurations result in piston/wave morphologies and kinematics that reproduce common characteristics of coronal waves. In the initial stage, the wave and the expanding source region cannot be clearly resolved; i.e. a certain time is needed before the wave detaches from the piston. Thereafter, it continues to travel as what is called a “simple wave.” During the acceleration stage of the source region inflation, the wave is driven by the piston expansion, so its amplitude and phase-speed increase, whereas the wavefront profile steepens. At a given point, a discontinuity forms in the wavefront profile; i.e. the leading edge of the wave becomes shocked. The time/distance required for the shock formation is shorter for a more impulsive source-region expansion. After the piston stops, the wave amplitude and phase speed start to decrease. During the expansion, most of the source region becomes strongly rarefied, which reproduces the coronal dimming left behind the eruption. However, the density increases at the source-region boundary, and stays enhanced even after the expansion stops, which might explain stationary brightenings that are sometimes observed at the edges of the erupted coronal structure. Also, in the rear of the wave a weak density depletion develops, trailing the wave, which is sometimes observed as weak transient coronal dimming. Finally, we find a well-defined relationship between the impulsiveness of the source-region expansion and the wave amplitude and phase speed. The results for the cylindrical piston are also compared with the outcome for a planar wave that is formed by a one-dimensional piston, to find out how different geometries affect the evolution of the wave.  相似文献   

In situ Raman observation of the crystallization in NaNO3–Na2SO4–H2O solution droplets     

Kirsten Linnow  Michael Steiger  Christine Lemster  Hilde De Clercq  Maja Jovanović 《Environmental Earth Sciences》2013,69(5):1609-1620

Several double salts have been detected in building materials and most of these salts are incongruently soluble compounds. In contrast to single salts, however, no systematic investigations of the crystallization behavior and deleterious effects of incongruently soluble double salts exist. To assess the damage potential of these salts, a systematic investigation of their highly complex behavior is desirable. This paper deals with the crystallization behavior of various solids in the ternary mixed NaNO₃–Na₂SO₄ system including the formation of the double salt darapskite, Na₃NO₃SO₄·H₂O. The crystallization sequence during droplet evaporation experiments at room conditions was determined using Raman and polarization microscopy. The basic idea of this research is to use deviations of the crystallization sequence of a salt or a mixed salt solution from the equilibrium pathway as an indicator to detect the degree of supersaturation. The observed crystallization pathway includes the formation of the metastable phases Na₂SO₄(III), Na₂SO₄(V) and darapskite. The experimental observations are discussed on the basis of the NaNO₃–Na₂SO₄–H₂O phase diagram and the results provide evidence for crystal growth from highly supersaturated solutions in both systems. If the crystals growing under these conditions are confined, these supersaturations result in substantial crystallization pressures.  相似文献