Thesis (Ph.D.)--Hong Kong University of Science and Technology, 2005. / Includes bibliographical references (leaves 125-134). Also available in electronic version.

Thesis--(Ph. D.)--University of Wisconsin--Madison, 1971. / Typescript. Vita. eContent provider-neutral record in process. Description based on print version record. Includes bibliography.

Thesis (MScIng)--University of Stellenbosch, 2005. / Bibliography. Also available via the Internet.

Thesis (Ph.D.)--Hong Kong University of Science and Technology, 2006. / Includes bibliographical references (leaves 143-165). Also available in electronic version.

Thesis (Ph. D.)--Oregon State University, 2008. / Printout. Includes bibliographical references (leaves 140-146). Also available on the World Wide Web.

No description available.

No description available.

The principal objective of this remote sensing and photogrammetric study was to extract information about the Chinese nuclear test site using commercial satellite imagery. As part of the research, image processing techniques were developed, applied, and refined. The extracted information was combined with other data - both technical and nontechnical - to increase knowledge of the test site and obtain a better understanding of the nuclear testing activities conducted at the site. The study consists of nine chapters. The first chapter provides the overall context, explains the research approach, and describes the scientific data that was selected for analysis. The second chapter critically reviews conceptual and empirical work on the remote sensing of nuclear test sites. The third chapter evaluates published work related to Chinese nuclear testing and the Chinese nuclear test site. The fourth chapter describes the capabilities and limitations of seismic sensing - the principal means of detecting underground nuclear explosions. The chapter also describes the early warning system that was developed to monitor recent seismic disturbances in the vicinity of the Chinese test site. This system proved useful in the ordering of satellite images of the Chinese test site. The fifth chapter describes the photogrammetric processing and analysis of Large Format Camera (LFC) stereo imagery of the test site. Augmenting tested software, new programs and scripts were written to generate a digital elevation model (DEM) and an orthoimage of the entire test site. Acquired shortly after an underground nuclear test in the area, the LFC imagery was also analyzed in order to detect, identify, and locate the detonation point as well as any other test site infrastructure. The sixth chapter details the analysis of three MSS images for surface effects related to atmospheric tests. Using temporal band ratioing, the images were processed to detect spectral and albedo variations caused by one multi-megaton nuclear test in the area. The seventh chapter contains the results from the use of image-derived information to reduce systematic errors in the seismic location estimates of Chinese tests. The refined location estimates revealed the existence of four distinct testing zones and provided insight on the role of each zone within the Chinese testing program. The refined locations were subsequently used to produce an orthoimage map of the test site. The eighth chapter describes the analysis of the underground testing zones using SPOT, Lands at TM, and ALMAZ. Using temporal band ratioing and the HRGB (Hue RGB) algorithm, the images were analyzed for surface changes that could be attributed to nuclear testing. Surface alterations were found and linked to nuclear testing activity, but none of the surface changes could be directly attributed to underground nuclear explosions. The final chapter summarizes the results, lists recommendations for further research, and details the conclusions of the study. The knowledge and experience that was obtained from this research can be used in future remote sensing and photogrammetric studies of nuclear testing. Key lessons were learned that could be applied to the monitoring of Chinese nuclear testing by the next generation of imaging satellites as well as to the verification of a future ban on nuclear testing.

In many machine learning applications data is assumed to be locally simple, where examples near each other have similar characteristics such as class labels or regression responses. Our goal is to exploit this assumption to construct locally simple yet globally complex systems that improve performance or reduce the cost of common machine learning tasks. To this end, we address three main problems: discovering and separating local non-linear structure in high-dimensional data, learning low-complexity local systems to improve performance of risk-based learning tasks, and exploiting local similarity to reduce the test-time cost of learning algorithms. First, we develop a structure-based similarity metric, where low-dimensional non-linear structure is captured by solving a non-linear, low-rank representation problem. We show that this problem can be kernelized, has a closed-form solution, naturally separates independent manifolds, and is robust to noise. Experimental results indicate that incorporating this structural similarity in well-studied problems such as clustering, anomaly detection, and classification improves performance. Next, we address the problem of local learning, where a partitioning function divides the feature space into regions where independent functions are applied. We focus on the problem of local linear classification using linear partitioning and local decision functions. Under an alternating minimization scheme, learning the partitioning functions can be reduced to solving a weighted supervised learning problem. We then present a novel reformulation that yields a globally convex surrogate, allowing for efficient, joint training of the partitioning functions and local classifiers. We then examine the problem of learning under test-time budgets, where acquiring sensors (features) for each example during test-time has a cost. Our goal is to partition the space into regions, with only a small subset of sensors needed in each region, reducing the average number of sensors required per example. Starting with a cascade structure and expanding to binary trees, we formulate this problem as an empirical risk minimization and construct an upper-bounding surrogate that allows for sequential decision functions to be trained jointly by solving a linear program. Finally, we present preliminary work extending the notion of test-time budgets to the problem of adaptive privacy.

No description available.