This thesis addresses the problem of deciding whether or not
two disjoint random samples that are fitted by the same regression
equation emanate from the same parental population. Two situations
are considered. The first situation is where the two disjoint samples
are individually fitted with the regression equation. The second...
A quantal response model, more general than the usual logistic
model, is introduced. This model takes into account sources of
variability, or experimental error, other than that arising from variability
in response between individual organisms (or other objects on
test). It is assumed that this extra source of variation is...
Anomaly detection is the task of identifying observations (points) that differ from the majority of other points, which requires some measure of difference, or distance. Many anomaly detection methods rely on “implicit distance” measures: rather than directly calculating an explicitly defined distance, these approaches quantify a point’s “abnormality” by examining...
Some nonparametric maximum likelihood estimation procedures
are developed for the class of pairs of distributions which have proportional
failure rate functions. Special consideration is given to the
case in which the shape of the failure rate functions are assumed to
be either increasing or decreasing. Estimators of the proportionality
constant,...
This thesis is concerned with the problem of developing a
method to categorize probability models by their outlier properties.
There have been two such categorization methods proposed in the
literature. Neyman and Scott (1971) classify an entire family of
distributions into the outlier properties outlier-prone completely
(OPC) and outlier resistant....
Differential expression (DE) analysis is a key task in gene expression study, because it uncovers the association between expression levels of a gene and the covariates of interest. This dissertation pertains to two particular aspects of DE analysis—identifying stably expressed genes for count normalization and accounting for correlation between DE...
RNA-Sequencing (RNA-Seq) has rapidly become the de facto technique in transcriptome studies. However, established statistical methods for analyzing experimental and observational microarray studies need to be revised or completely re-invented to accommodate RNA-Seq data's unique characteristics. In this dissertation, we focus on statistical analyses performed at two particular stages in...
The recent advent of large-scale microbiome studies enabled by high-throughput sequencing calls for innovative statistical methodologies that are capable of tackling a myriad of challenges presented by microbiome data. Compelled by this need, we focus on the development of statistical tools for two types of problems that arise in microbiome...
When data are not missing at random, approaches to reduce nonresponse bias include subsampling nonresponding units and modeling. The objective of this thesis is to develop unbiased and precise model-assisted estimators of the population total that are applicable to data from a complex survey design with nonignorable nonresponse. When information...
.......................................................................................... 104
Table 4.5: Pearson chi-square tests of association for indicators of response and
success