Physical Sciences and Mathematics Commons^™

Convergence Rates For Empirical Estimation Of Binary Classification Bounds, Salimeh Yasaei Sekeh, Morteza Noshad, Kevin R. Moon, Alfred O. Hero

Mathematics and Statistics Faculty Publications

Bounding the best achievable error probability for binary classification problems is relevant to many applications including machine learning, signal processing, and information theory. Many bounds on the Bayes binary classification error rate depend on information divergences between the pair of class distributions. Recently, the Henze–Penrose (HP) divergence has been proposed for bounding classification error probability. We consider the problem of empirically estimating the HP-divergence from random samples. We derive a bound on the convergence rate for the Friedman–Rafsky (FR) estimator of the HP-divergence, which is related to a multivariate runs statistic for testing between two distributions. The FR estimator is …

Extensions And Improvements To Random Forests For Classification, Anna Quach

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

The motivation of my dissertation is to improve two weaknesses of Random Forests. One, the failure to detect genetic interactions between two single nucleotide polymorphisms (SNPs) in higher dimensions when the interacting SNPs both have weak main effects and two, the difficulty of interpretation in comparison to parametric methods such as logistic regression, linear discriminant analysis, and linear regression.

We focus on detecting pairwise SNP interactions in genome case-control studies. We determine the best parameter settings to optimize the detection of SNP interactions and improve the efficiency of Random Forests and present an efficient filtering method. The filtering method is …

Real Simple Lie Algebras: Cartan Subalgebras, Cayley Transforms, And Classification, Hannah M. Lewis

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

The differential geometry software package in Maple has the necessary tools and commands to automate the classification process for complex simple Lie algebras. The purpose of this thesis is to write the programs to complete the classification for real simple Lie algebras. This classification is difficult because the Cartan subalgebras are not all conjugate as they are in the complex case. For the process of the real classification, one must first identify a maximally noncompact Cartan subalgebra. The process of the Cayley transform is used to find this specific Cartan subalgebra. This Cartan subalgebra is used to find the simple …

Classification Of Five-Dimensional Lie Algebras With One-Dimensional Subalgebras Acting As Subalgebras Of The Lorentz Algebra, Jordan Rozum

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

Motivated by A. Z. Petrov's classification of four-dimensional Lorentzian metrics, we provide an algebraic classification of the isometry-isotropy pairs of four-dimensional pseudo-Riemannian metrics admitting local slices with five-dimensional isometries contained in the Lorentz algebra. A purely Lie algebraic approach is applied with emphasis on the use of Lie theoretic invariants to distinguish invariant algebra-subalgebra pairs. This method yields an algorithm for identifying isometry-isotropy pairs subject to the aforementioned constraints.

The Classification Of Simple Lie Algebras In Maple, D. Russell Sadler

All Graduate Plan B and other Reports, Spring 1920 to Spring 2023

Lie algebras are invaluable tools in mathematics and physics as they enable us to study certain geometric objects such as Lie groups and differentiable manifolds. The computer algebra system Maple has several tools in its Lie Algebras package to work with Lie algebras and Lie groups. The purpose of this paper is to supplement the existing software with tools that are essential for the classification of simple Lie algebras over C.

In particular, we use a method to find a Cartan subalgebra of a Lie algebra in polynomial time. From the Cartan subalgebra we can compute the corresponding root system. …

Comparison Of Machine Learning Algorithms For Modeling Species Distributions: Application To Stream Invertebrates From Western Usa Reference Sites, Margi Dubal

All Graduate Plan B and other Reports, Spring 1920 to Spring 2023

Machine learning algorithms are increasingly being used by ecologists to model and predict the distributions of individual species and entire assemblages of sites. Accurate prediction of distribution of species is an important factor in any modeling. We compared prediction accuracy of four machine learning algorithms-random forests, classification trees, support vector machines, and gradient boosting machines to a traditional method, linear discriminant models (LDM), on a large set of stream invertebrate data collected at 728 reference sites in the western United States. Classifications were constructed for individual species and for assemblages of sites clustered a priori by similarity on biological characteristics. …

A Classification Of Real Indecomposable Solvable Lie Algebras Of Small Dimension With Codimension One Nilradicals, Alan R. Parry

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

This thesis was concerned with classifying the real indecomposable solvable Lie algebras with codimension one nilradicals of dimensions two through seven. This thesis was organized into three chapters.

In the first, we described the necessary concepts and definitions about Lie algebras as well as a few helpful theorems that are necessary to understand the project. We also reviewed many concepts from linear algebra that are essential to the research.

The second chapter was occupied with a description of how we went about classifying the Lie algebras. In particular, it outlined the basic premise of the classification: that we can use …

Special Classification Models For Lichens In The Pacific Northwest, Janeen Ardito

All Graduate Plan B and other Reports, Spring 1920 to Spring 2023

A common problem in ecological studies is that of determining where to look for rare species. This paper shows how statistical models, such as classification trees, may be used to assist in the design of probability-based surveys for rare species using information on more abundant species that are associated with the rare species. This model assisted approach to survey design involves first building models for the more abundant species. The models are then used to determine stratifications for the rare species that are associated with the more abundant species. The goal of this approach is to increase the number of …

Lorentz Homogeneous Spaces And The Petrov Classification, Adam Bowers

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

A. Z. Petrov gave a complete list of all local group actions on a four-dimensional space-time that admit an invariant Lorentz metric, up to an equivalence relation. His list was compiled by directly constructing all possible Lie algebras of infinitesimal generators of group actions that preserve a Lorentz metric. The goal of this paper was to verify that classification by algebraically constructing a list of all possible three-dimensional homogeneous spaces and calculating which among them have a non-degenerate invariant metric.

The Classification Of Low Dimensional Nilpotent Lie Algebras, Kimberli C. Tripp

All Graduate Plan B and other Reports, Spring 1920 to Spring 2023

Nilpotent Lie algebras are the fundamental building blocks for generic (not semi-simple) Lie algebras. In particular, the classification of nilpotent algebras is the first step in classifying and identifying solvable Lie Algebras. The problem of classifying nilpotent Lie algebras was first studied by Umlauf [9] in 1891. More recently, classifications have been given up to dimension six using different techniques by Morosov (1958) [7], Skjelbred and Sund (1977) [8], and up to dimension five by Dixmier (1958) [2]. Using Morosov's method of classification by maximal abelian ideals, Winternitz reproduced the Morosov classification obtaining different canonical forms for the algebras. The …

A New Perspective On Classification, Guohua Zhao

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

The idea of voting multiple decision rules was introduced in to statistics by Breiman. He used bootstrap samples to build different decision rules, and then aggregated them by majority voting (bagging). In regression, bagging gives improved predictors by reducing the variance (random variation), while keeping the bias (systematic error) the same. Breiman introduced the idea of bias and variance for classification to explain how bagging works. However, Friedman showed that for the two-class situation, bias and variance influence the classification error in a very different way than they do in the regression case.

In the first part of …