Master of Science (MS), Wright State University, 2011, Biological Sciences

My study was conducted to understand tree community structure and how soil moisture and fire frequency influence them. Eighteen plots were placed in the Edge of Appalachia Nature Preserve of unglaciated southern Ohio: nine within a prescribed burn site and nine control sites outside the burn. Sites were stratified in triplicate across GIS-derived integrated soil moisture index (IMI) classes. Burning was done in 1996. Overstory species dbh and sapling species were sampled 1997, 2001, and 2008. Overstory stems were located in 2009 using range finders. Stem locations were loaded into GIS using novel techniques to quantify individual stem IMI values. Nonmetric multi-dimensional scaling identified greater heterogeneity among intermediate and mesic sites than xeric sites. Multi-response permutation procedures did not detect community differences between burned and unburned sites, but did detect strong (A=0.3 to 0.2, T=-3.6 to -4.1) distinct community differences that were statistically significant (P < 0.05) among xeric, intermediate, and mesic IMI classes. Analysis of variance identified significant initial effects of burning on Carya saplings and overstory Sassafras albidum stems, as well as lasting effects significant on Carpinus caroliniana. ANOVA detected significant differences across all sampling years in sapling relative number for Acer rubrum, Sassafras albidum, and Carpinus caroliniana saplings, as well as Quercus prinus, and Liriodendron tulipifera overstory stems between IMI classes. Bonferroni adjusted Kolmogorov-Smirnov tests were used to identify and quantify IMI habitat restrictions of species. Quercus prinus dominated xeric sites (IMI quartiles 18-24), Carya occupied intermediate sites (IMI quartiles 22-44), Acer saccharum occupied intermediate to mesic sites (IMI quartiles 33-44), Sassafras albidum (IMI quartiles 20-40, IMI median 43) and Liriodendron tulipifera (IMI quartiles 39-45) were restricted to mesic sites. My r (open full item for complete abstract)

Committee: James Runkle PhD (Advisor); James Amon PhD (Committee Member); Thomas Rooney PhD (Committee Member) Subjects: Ecology

MS, Kent State University, 2012, College of Arts and Sciences / Department of Computer Science

Understanding the differences between two datasets is a fundamental data mining question and is also ubiquitously important across many real world scientific applications. In this paper, we propose a tree-based framework to provide a parsimonious explanation of the difference between two distributions based on rigorous two-sample statistical test. We develop two efficient approaches. The first one is a dynamic programming approach that finds a minimal number of data subsets that describe the difference between two data sets. The second one is a greedy approach that approximates the dynamic programming approach. We employ the well-known Friedman's MST (minimal spanning tree) statistics for two-sample statistical tests in our summarization tree construction, and develop novel techniques to speedup its computational procedure. We performed a detailed experimental evaluation on both real and synthetic datasets and demonstrated the effectiveness of our tree-summarization approach.

Committee: Ruoming Jin (Advisor); Yuri Breitbart (Advisor); Feodor Dragan (Committee Member); Peyravi Hassan (Committee Member) Subjects: