Home
Search results “Cressie statistics for spatial data mining”
Spatio-Temporal Statistics in Geodesign
 
50:23
Distinguished professor Dr. Noel Cressie from the University of Wollongong brings his award-winning studies in spatial statistics to geodesign in this thought-provoking keynote. Cressie shows how introducing conditional probability to geodesign allows spatio-temporal statistics to handle uncertainties in data.
Views: 1157 Esri Geodesign
An interview with Noel Cressie
 
08:06
An interview with Distinguished Professor Noel Cressie of the University of Wollongong, a big name in spatial statistics, advocate of hierarchical modeling in ecology, and author of a key reference text in spatial statistics, and more recently "Statistics for Spatio-temporal data" with Chris Wikle; we discuss all of these topics.
Views: 1114 MethodsEcolEvol
Discussion: Spatio-Temporal Statistics
 
21:51
Noel Cressie, Jack Dangermond, and Carl Steinitz join Tom Fisher to discuss creating policy through geodesign at both large and small scales using the Cressie's principles of conditional probability.
Views: 341 Esri Geodesign
Data Issues: Multiple Testing, Bias, Confounding, Missing...
 
01:00:59
Dr. Lance Waller from Emory University presents a lecture titled "Data Issues: Multiple Testing, Bias, Confounding, & Missing Data." View Slides https://drive.google.com/open?id=0B4IAKVDZz_JUczRSd0NucjlhT00 Lecture Abstract Once data are scraped, wrangled, linked, merged, and analyzed, what information do they reveal and can we trust the resulting conclusions? In this presentation, we define and review data issues relating to the analysis and interpretation of observational data from the field of epidemiology and consider implications for data science, especially regarding the goal of moving from big data to knowledge. Specifically, we explore concepts of bias, confounding, effect modification, and missing/mismeasured data as applied to data science. We provide an analytic context based on sampling concepts and explore relevant literature and tools from epidemiology, biostatistics, computer science, and data science. As with many issues in data science, the full applicability of the concepts is very much a work in progress and present multiple opportunities for future development. About the Speaker Lance A. Waller, Ph.D. is Rollins Professor and Chair of the Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University. He is a member of the National Academy of Science Committee on Applied and Theoretical Statistics. His research involves the development of statistical methods for geographic data including applications in environmental justice, epidemiology, disease surveillance, spatial cluster detection, conservation biology, and disease ecology. His research appears in biostatistical, statistical, environmental health, and ecology journals and in the textbook Applied Spatial Statistics for Public Health Data (2004, Wiley). Join our weekly meetings from your computer, tablet or smartphone. Visit our website to view our schedule and join our next live webinar! http://www.bigdatau.org/data-science-seminars
CAM Colloquium: Christopher Wikle
 
53:44
Friday, April 18, 2014 An Overview of Mechanistically-Motivated Dynamic Spatio-Temporal Statistical Models Spatio-temporal statistical models are increasingly being used across a wide variety of scientific disciplines to describe and predict spatially-explicit processes that evolve over time. Correspondingly, in recent years there has been a significant amount of research on new statistical methodology for such models. Although descriptive models that approach the problem from the second-order (covariance) perspective are important, and innovative work is being done in this regard, many real-world processes are dynamic, and it can be more efficient in some cases to characterize the associated spatio-temporal dependence by the use of dynamic models. The chief challenge with the specification of such dynamical models has been related to the curse of dimensionality and the specification of realistic dependence. Even in fairly simple linear, first-order Markovian settings with Gaussian errors, statistical models are often over parameterized. Hierarchical models have proven invaluable in their ability to deal to some extent with this issue by allowing dependency among groups of parameters. In addition, this framework has allowed for the specification of science-based parameterizations (and associated prior distributions) in which classes of mechanistic dynamical models (e.g., partial differential equations (PDEs), integro-difference equations (IDEs), and agent-based models (ABMs)) are used to motivate specific parameterizations. Most of the focus for the application of such models in statistics has been in the linear case. The problems mentioned above with linear dynamic models are compounded in the case of nonlinear models, yet these are the processes that govern environmental science. In this sense, the need for coherent and sensible model parameterizations is not only helpful, it is essential. Here, we present on overview of recent research for accommodating realistic “science-based” linear and nonlinear motivating structure in spatio-temporal dynamical models as well as some discussion of dimension reduction, model selection and the use of emulators or model surrogates for prior elicitation. Examples will be presented from atmospheric science, oceanography, and ecology.
Kriging
 
25:16
In statistics, originally in geostatistics, Kriging or Gaussian process regression is a method of interpolation for which the interpolated values are modeled by a Gaussian process governed by prior covariances, as opposed to a piecewise-polynomial spline chosen to optimize smoothness of the fitted values. Under suitable assumptions on the priors, Kriging gives the best linear unbiased prediction of the intermediate values. Interpolating methods based on other criteria such as smoothness need not yield the most likely intermediate values. The method is widely used in the domain of spatial analysis and computer experiments. The technique is also known as Kolmogorov Wiener prediction. The theoretical basis for the method was developed by the French mathematician Georges Matheron based on the Master's thesis of Danie G. Krige, the pioneering plotter of distance-weighted average gold grades at the Witwatersrand reef complex in South Africa. Krige sought to estimate the most likely distribution of gold based on samples from a few boreholes. The English verb is to krige and the most common noun is kriging; both are often pronounced with a hard "g", following the pronunciation of the name "Krige". This video is targeted to blind users. Attribution: Article text available under CC-BY-SA Creative Commons image source in video
Views: 11069 Audiopedia