Thesis (Professional Degree)--University of Missouri, School of Mines and Metallurgy, 1930. / The entire thesis text is included in file. Typescript. Illustrated by author. Title from title screen of thesis/dissertation PDF file (viewed November 30, 2009) Includes bibliographical references (p. 26).

Thesis (M.S.)--West Virginia University, 2002. / Title from document title page. Document formatted into pages; contains x, 135 p. : ill. (some col.), col. maps. Includes abstract. Includes bibliographical references (p. 134-135).

The Texas Department of Transportation (TxDOT) has sought to reduce expenditures by better utilizing their maintenance forces through a compressed workweek. The focus of this thesis is a before and after comparison of maintenance crews at TxDOT during a standard 5-day forty hour workweek and a compressed 4-day forty hour workweek. Compressed workweeks are work arrangements in which full-time employees are allowed to work longer days for part of the week or pay period in exchange for shorter days or a day off during the same week or pay period. This type of schedule allows for numerous benefits including increased productivity, additional time to handle personal business, less travel time, less start up and shut down time, improved morale, as well as less stress. Originally three districts were placed upon the compressed workweek but after four months of trial, three additional districts were included. Maintenance activity data from previous years was compared to data collected over the trial period to assess productivity impacts as well as vehicular travel. Surveys of maintenance crews were conducted throughout the study to address personal concerns. The results of the study were that there were no significant impacts to productivity or vehicular usage but a significant improvement in work conditions for the maintenance crews. Further assessment is recommended utilizing additional functional codes for more illustrative results. / text

No description available.

Compressed Sensing (CS) is a framework where we measure data through a non-adaptive linear mapping with far fewer measurements that the ambient dimension of the data. This is made possible by the exploitation of the inherent structure (simplicity) in the data being measured. The central issues in this framework is the design and analysis of the measurement operator (matrix) and recovery algorithms. Restricted isometry constants (RIC) of the measurement matrix are the most widely used tool for the analysis of CS recovery algorithms. The addition of the subscripts 1 and 2 below reflects the two RIC variants developed in the CS literature, they refer to the ℓ1-norm and ℓ2-norm respectively. The RIC2 of a matrix A measures how close to an isometry is the action of A on vectors with few nonzero entries, measured in the ℓ2-norm. This, and related quantities, provide a mechanism by which standard eigen-analysis can be applied to topics relying on sparsity. Specifically, the upper and lower RIC2 of a matrix A of size n × N is the maximum and the minimum deviation from unity (one) of the largest and smallest, respectively, square of singular values of all (N/k)matrices formed by taking k columns from A. Calculation of the RIC2 is intractable for most matrices due to its combinatorial nature; however, many random matrices typically have bounded RIC2 in some range of problem sizes (k, n,N). We provide the best known bound on the RIC2 for Gaussian matrices, which is also the smallest known bound on the RIC2 for any large rectangular matrix. Our results are built on the prior bounds of Blanchard, Cartis, and Tanner in Compressed Sensing: How sharp is the Restricted Isometry Property?, with improvements achieved by grouping submatrices that share a substantial number of columns. RIC2 bounds have been presented for a variety of random matrices, matrix dimensions and sparsity ranges. We provide explicit formulae for RIC2 bounds, of n × N Gaussian matrices with sparsity k, in three settings: a) n/N fixed and k/n approaching zero, b) k/n fixed and n/N approaching zero, and c) n/N approaching zero with k/n decaying inverse logarithmically in N/n; in these three settings the RICs a) decay to zero, b) become unbounded (or approach inherent bounds), and c) approach a non-zero constant. Implications of these results for RIC2 based analysis of CS algorithms are presented. The RIC2 of sparse mean zero random matrices can be bounded by using concentration bounds of Gaussian matrices. However, this RIC2 approach does not capture the benefits of the sparse matrices, and in so doing gives pessimistic bounds. RIC1 is a variant of RIC2 where the nearness to an isometry is measured in the ℓ1-norm, which is both able to better capture the structure of sparse matrices and allows for the analysis of non-mean zero matrices. We consider a probabilistic construction of sparse random matrices where each column has a fixed number of non-zeros whose row indices are drawn uniformly at random. These matrices have a one-to-one correspondence with the adjacency matrices of fixed left degree expander graphs. We present formulae for the expected cardinality of the set of neighbours for these graphs, and present a tail bound on the probability that this cardinality will be less than the expected value. Deducible from this bound is a similar bound for the expansion of the graph which is of interest in many applications. These bounds are derived through a more detailed analysis of collisions in unions of sets using a dyadic splitting technique. This bound allows for quantitative sampling theorems on existence of expander graphs and the sparse random matrices we consider and also quantitative CS sampling theorems when using sparse non mean-zero measurement matrices.

In this dissertation we address the problem of adapting frequency domain tiling using the curvelet transform as the basis algorithm. The optimal tiling, for a given class of images, is computed using denoising performance as the cost function. The major adaptations considered are: the number of scale decompositions, angular decompositions per scale/quadrant, and scale locations. A global optimization algorithm combining the three adaptations is proposed. Denoising performance of adaptive curvelets is tested on seismic and face data sets. The developed adaptation procedure is applied to a number of different application areas. Adaptive curvelets are used to solve the problem of sparse data recovery from subsampled measurements. Performance comparison with default curvelets demonstrates the effectiveness of the adaptation scheme. Adaptive curvelets are also used in the development of a novel image similarity index. The developed measure succeeds in retrieving correct matches from a variety of textured materials. Furthermore, we present an algorithm for classifying different types of seismic activities.

Thesis (Ph.D.)--Ohio State University, 1999. / Includes bibliographical references (leaves 184-196).

Thesis (Ph. D.)--Lehigh University, 1998. / Includes vita. Bibliography: leaves 102-110.

Thesis (M.S in Transportation Tecnology and Policy)--University of California, Davis, 2001. / "October 2001." Includes bibliographical references (p. 85-87). Also available online via the ITS Davis website (its.ucdavis.edu).

Thesis (Ph. D.)--West Virginia University, 2009. / Title from document title page. Document formatted into pages; contains viii, 162 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 138-142).