Date of Award

Spring 5-21-2026

Document Type

Honors Project

University Scholars Director

Dr. Joshua Tom

First Advisor/Committee Member

Dr. Carlos R. Arias

Keywords

machine learning, natural language processing, gated recurrent neural networks, empirical evidence, classification

Abstract

Gated recurrent neural networks such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU) help fix instability present in normal recurrent neural networks. This allows them to be used for various real-world tasks, and due to their architecture, they are uniquely qualified to handle variable sized input such as text. However, even before training can begin on a machine learning model, various hyperparameters must be chosen to decide how the model will be architectured. Choosing good hyperparameters is vital for creating a model that performs well but is not larger and more computationally expensive to run than it needs to be. This paper identifies various hyperparameters in both LSTM and GRU models and empirically tests a wide range of hyperparameter values across various different natural language processing (NLP) tasks and finds the resulting accuracy. From there, various graphs are presented and patterns are identified which can highlight various properties of each hyperparameter. Using these results, the developer is able to create more informed decisions when deciding what hyperparameter values to choose when creating a gated recurrent neural network from scratch.

Comments

A project submitted in partial fulfillment of the requirements of the University Scholars Honors Program.

Copyright Status

http://rightsstatements.org/vocab/InC/1.0/

Additional Rights Information

Copyright held by author.

Download

Share

COinS
 
Copyright Status