scholarly journals Bayesian inference for big spatial data using non-stationary spectral simulation

2021 ◽  
pp. 100507
Author(s):  
Hou-Cheng Yang ◽  
Jonathan R. Bradley
Keyword(s):  
Bayesian Inference ◽  
Spatial Data ◽  
Spectral Simulation

scholarly journals Efficient Bayesian inference for large chaotic dynamical systems

2021 ◽  
Vol 14 (7) ◽  
pp. 4319-4333
Author(s):  
Sebastian Springer ◽  
Heikki Haario ◽  
Jouni Susiluoto ◽  
Aleksandr Bibov ◽  
Andrew Davis ◽  
...  
Keyword(s):  
Bayesian Inference ◽  
Dynamical Systems ◽  
Spatial Data ◽  
Large Scale ◽  
Likelihood Function ◽  
Computational Cost ◽  
Local Approximation ◽  
Physical Models ◽  
Chaotic Dynamical Systems ◽  
Time Required

Abstract. Estimating parameters of chaotic geophysical models is challenging due to their inherent unpredictability. These models cannot be calibrated with standard least squares or filtering methods if observations are temporally sparse. Obvious remedies, such as averaging over temporal and spatial data to characterize the mean behavior, do not capture the subtleties of the underlying dynamics. We perform Bayesian inference of parameters in high-dimensional and computationally demanding chaotic dynamical systems by combining two approaches: (i) measuring model–data mismatch by comparing chaotic attractors and (ii) mitigating the computational cost of inference by using surrogate models. Specifically, we construct a likelihood function suited to chaotic models by evaluating a distribution over distances between points in the phase space; this distribution defines a summary statistic that depends on the geometry of the attractor, rather than on pointwise matching of trajectories. This statistic is computationally expensive to simulate, compounding the usual challenges of Bayesian computation with physical models. Thus, we develop an inexpensive surrogate for the log likelihood with the local approximation Markov chain Monte Carlo method, which in our simulations reduces the time required for accurate inference by orders of magnitude. We investigate the behavior of the resulting algorithm with two smaller-scale problems and then use a quasi-geostrophic model to demonstrate its large-scale application.

scholarly journals Efficient Bayesian inference for large chaotic dynamical systems

2020 ◽  
Author(s):  
Sebastian Springer ◽  
Heikki Haario ◽  
Jouni Susiluoto ◽  
Aleksandr Bibov ◽  
Andrew Davis ◽  
...  
Keyword(s):  
Bayesian Inference ◽  
Dynamical Systems ◽  
Spatial Data ◽  
Large Scale ◽  
Likelihood Function ◽  
Computational Cost ◽  
Local Approximation ◽  
Physical Models ◽  
Chaotic Dynamical Systems ◽  
Time Required

Abstract. Estimating parameters of chaotic geophysical models is challenging due to these models' inherent unpredictability. With temporally sparse long-range observations, these models cannot be calibrated using standard least squares or filtering methods. Obvious remedies, such as averaging over temporal and spatial data to characterize the mean behavior, do not capture the subtleties of the underlying dynamics. We perform Bayesian inference of parameters in high-dimensional and computationally demanding chaotic dynamical systems by combining two approaches: (i) measuring model-data mismatch by comparing chaotic attractors, and (ii) mitigating the computational cost of inference by using surrogate models. Specifically, we construct a likelihood function suited to chaotic models by evaluating a distribution over distances between points in the phase space; this distribution defines a summary statistic that depends on the attractor's geometry, rather than on pointwise matching of trajectories. This statistic is computationally expensive to simulate, compounding the usual challenges of Bayesian computation with physical models. Thus we develop an inexpensive surrogate for the log-likelihood via local approximation Markov chain Monte Carlo, which in our simulations reduces the time required for accurate inference by orders of magnitude. We investigate the behavior of the resulting algorithm on model problems, and then use a quasi-geostrophic model to demonstrate its large-scale application.

scholarly journals Bayesian Inference of Ecological Interactions from Spatial Data

Entropy ◽  
2017 ◽  
Vol 19 (12) ◽  
pp. 547 ◽  
Author(s):  
Christopher Stephens ◽  
Victor Sánchez-Cordero ◽  
Constantino González Salazar
Keyword(s):  
Bayesian Inference ◽  
Spatial Data ◽  
Ecological Interactions

ANALYSIS OF FLOOD PRONE AREAS IN NAGAN RAYA REGENCY ACEH PROVINCE

2020 ◽  
Vol 5 (1) ◽  
pp. 414
Author(s):  
Amsar Yunan
Keyword(s):  
Land Use ◽  
Spatial Data ◽  
Flood Hazard ◽  
Parameter Analysis ◽  
Thematic Maps ◽  
Aceh Province ◽  
Object Area ◽  
Use Factors ◽  
Hazard Levels ◽  
Flood Prone Areas

Maps or remote sensing can be interpreted as the process of reading using various sensors where data collected remotely can be analyzed to obtain information about the object, area or phenomenon. In this study, the author develops a flood disaster mapping information system applying overlays with scoring between the parameters. The determinant factors to provide flood hazard levels includes rainfall factors in the dasarian unit, land-use factors and land-use arbitrary factors. Of all these parameters, a scoring process will be carried out by assigning weights and values according to their respective classifications, then an overlay process will be performed using ArcGIS software. The author conducted this study in Nagan Raya Regency since this area experiences flooding annually.  Framing a thematic map of flood-prone areas in Nagan Raya Regency was designed using the flood hazard method. Spatial data that has been presented in the form of thematic maps as parameters are land use maps, landform maps, and dasarian rainfall maps (per 10 daily). The design of thematic maps that are prone to flooding is done by overlapping (overlay process). In contrast, the determination of the classification is done by adding scores to each parameter, with low, medium and high hazard levels. Parameter analysis shows the level of flood vulnerability in Nagan Raya Regency of each district, namely Beutong: high 0.21%, medium 13.68%, low 86.12%. Seunagan District: high 51.17%, medium 48.83%, low 0%. Seunagan Timur District: high 10.07%, medium 46.18%, low 43.75%. Kuala Subdistrict: high 29.66%, medium 68.99%, low 1.35%. Darul Makmur District: high 8.57%, medium 63.37%, low 28.06%. From the overall results of the study, it can be concluded that the danger of flooding in Nagan Raya Regency with a level of vulnerability: high 9.92%, moderate 42.65% and low 47.43%.

An empirical analysis of town-forming enterprises' corporate social responsibility: Evidence from the Volga Federal District

2020 ◽  
Vol 18 (10) ◽  
pp. 1894-1909
Author(s):  
I.R. Badykova
Keyword(s):  
Regression Analysis ◽  
Corporate Social Responsibility ◽  
Social Responsibility ◽  
Empirical Analysis ◽  
Spatial Data ◽  
Univariate Analysis ◽  
Federal District ◽  
Current Situation ◽  
Economic Situation ◽  
Corporate Social

Subject. This article explores the determinants of social responsibility of backbone enterprises. Objectives. The article aims to investigate the relationships between the socio-economic situation of the monotown where the backbone company operates, and corporate social responsibility (CSR). Methods. For the study, I used a regression analysis and univariate analysis of spatial data. The rating estimates calculated using an original methodology are used as a CSR proxy (dependent variable). Results. Presenting information about the current situation of backbone enterprises and monotowns in Russia, the article reveals the existence of relationships between the backbone enterprise's affiliation to a monotown with a certain socio-economic situation and the level of corporate social responsibility. Conclusions. The situation of the backbone companies is likely to deteriorate. Increasing the level of social responsibility during a crisis seems unlikely.

Sign in / Sign up

Export Citation Format

Share Document