Back to search

Cornell University

Spatial and Temporal Approaches to Analyzing Big Data

Abstract

dc:description.abstract

Modern technology allows researchers across disciplines to capture vast quantities of data -- ``big" data -- to help answer some of their most pressing scientific questions. The analysis of big data, though, is complicated by its computational feasibility and the availability of appropriate statistical techniques. These problems only grow in complexity when working with dependent data, such as data which is recorded across time (temporal dependence), or data which is recorded across space (spatial dependence). In this work, we offer tools to aid analysis of big data in both domains. In the temporal domain, we study a classification approach for high-dimensional times series data. The motivating dataset for this work comes from biological time series data simultaneously measured across many electrical frequencies. Our results suggest the plausibility of accurately classifying a variety of cell lines, thus providing researchers with a means of ``checking" the cell types under study prior to reporting any results. In the spatial domain, we present work for fitting statistical models to large spatial point pattern data. In particular, we combine spectral and Laplace approximations, among others, in an EM algorithm, denoted SLEM, to approximately fit the widely popular log-Gaussian Cox process to these big spatial datasets. Given the utility of SLEM, we are able to conduct a large-scale lightning dynamics study across the contiguous United States, where we make interesting observations about the relationship between lightning occurrence and several environmental covariates.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Statistics
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Statistics
Grantor
Cornell University
Year dc:date.issued
2021

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Gelsinger, Megan Lynne
Chair dc:contributor.chair
  • Matteson, David
Committee members dc:contributor.committeemember
  • Guinness, Joe
  • Booth, James
  • Basu, Sumanta

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
15160176
Dc Identifier Other
ProQuest Submission ID: 12624
ProQuest Publication ID: 28648234

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Gelsinger, Megan Lynne. Spatial and Temporal Approaches to Analyzing Big Data. Doctor of Philosophy thesis, Cornell University, 2021. https://hdl.handle.net/1813/110548