Georgia Southern University
Time Series Analysis Using Hierarchical Neural Network-Based Models
Abstract
dc:description.abstractTime series are numerical sequences typically produced by collecting process data at regular time intervals. They appear in many applications and their analysis has been a continuous research topic over many decades. Artificial Neural Networks are sophisticated mathematical models inspired by how human brain neurons are interconnected to produce decisions and judgements. They are parameter models that can be adapted to successfully discover trends and patterns in various types of static or dynamic data (sequential, images, sound, etc.), and produce predictions of further behavior of dynamic data. This work adapts the power of the Artificial Neural Networks and proposes a novel hierarchical model for finding trends and cyclic patterns in time series data. Our model can be used to perform practical analysis of past time series data (financial, sales, or biological data) and can provide useful insights into the near future evolution of such data.
Degree
thesis:*- Name thesis:degree_name
- Master of Science in Mathematics (M.S.)
- Level thesis:degree_level
- Thesis (restricted to Georgia Southern)
- Discipline thesis:degree_discipline
- Department of Mathematical Sciences
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Farotimi, Oluwasayo
- Contributors dc:contributor
-
- Stephen Carden
- Zheni Utic
Subjects
dc:subject × 8Rights
- Licence dc:rights.license
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://digitalcommons.georgiasouthern.edu/etd/2744