- Shopping Bag ( 0 items )
Ships from: Columbus, OH
Usually ships in 1-2 business days
Ships from: Dallas, TX
Usually ships in 1-2 business days
|Ch. 2||Univariate Data Analysis||9|
|Ch. 3||Bivariate Data Analysis||32|
|Ch. 4||Multivariate Data Analysis||61|
|Ch. 5||Univariate Spatial Analysis||85|
|Ch. 6||Multivariate Spatial Data Analysis||132|
|Ch. 7||Spatial Simulation||159|
|Ch. 8||Digital Image Processing||180|
|Ch. 9||Composite Visualizations||219|
|A||Critical Values of the Chi-Square Distribution||248|
|B||Critical Values of Squared Correlation Coefficient, p-plot||251|
|C||Critical Values of F Distribution||253|
|D||Critical Values of t Distribution||257|
Readers will learn about statistics that describe data. From this fundamental foundation, more sophisticated algorithms for analyzing and manipulating data are applied. First, regression analysis is described and shown to yield a metric for correlation. This metric is of paramount importance to multivariate analysis. Moreover, space is of critical concern to many scientific and engineering applications, consequently applications in spatial analysis, both univariate and multivariate, are described in which correlation is, again, the metric of paramount importance. Spatial simulations are then shown to yield visually exciting images that enable experiments with spatial data to enhance the understanding of their behavior. A widely utilized form of spatial data is that which is obtained from digital sensors, spaceborne sensors in particular. Methods of data analysis described in the text are applied to these data to yield visualizations that aid a reader's understanding of them. Finally, composite visualizations are introduced, such as the draping of Landsat TM data over digital elevation to understand how image features correlate with elevation. Such composite visualizations are also used to associatecolor with correlation. In general, the unifying theme of this text is the analysis and understanding of correlation.
Following an introductory chapter that motivates the use and understanding of the text, chapters are presented in order on univariate data analysis, bivariate data analysis (regression and correlation), multivariate data analysis, univariate spatial analysis, multivariate spatial analysis, spatial simulation (including experiments with random number generation), digital image processing, and composite visualizations. A CD is included at the back of the text that presents a program, Visual_Data, a Windows-based program capable of reproducing all analytical results shown in the text. Moreover, the CD contains all data sets and digital images shown in the text, as well as many that are not shown to enable readers to experiment with data analysis beyond what is shown in the text. Readers are not forced to use Visual_Data. Instead, this program is provided should other software not be available. The text presents many examples of data analysis using Microsoft's Excel spreadsheet software and MATLAB, two powerful, commercial software platforms should readers prefer their use.
This book is written from the perspective of eighteen years of college-level instruction on data analysis in geological engineering and geology programs. This experience has shown that too much emphasis on mathematics when describing algorithms leaves too little time for a discussion of applications. But, a superficial treatment of mathematics trivializes algorithms, risking their use as "black boxes." This text was written to attempt a balance between the mathematical description of algorithms and visualizations of their outcomes. Moreover, these visualizations provide useful insights to algorithms that are not necessarily obvious in their equation form. Most importantly, visualizations of data are emphasized. A single data set, a collection of Landsat TM spectral bands, is used throughout this text as a unifying theme, showing how seemingly independent algorithms for data analysis are, in fact, interrelated, moreover showing that useful analyses of data can be obtained even if they don't conform to some ideal notion of statistical analysis, such as conformity to the normal distribution. A fundamental aspect of this text is on visualizing the data themselves, thinking critically of all outcomes from their analysis to judge whether useful insight to their behavior has been achieved.