Physical Sciences and Mathematics Commons^™

Super Mario Evolution By The Augmentation Of Topology, Russell A. Autin

University of New Orleans Theses and Dissertations

This paper describes the creation and development of an implementation of the NeuroEvolution of Augmenting Topologies (NEAT) architecture to train an agent to play Super Mario Brothers. Building off of a basic implementation of NEAT, this thesis project shows the process of refining the fitness calculation that ranks the networks in the population and also defines the creation and application of a dataset to train the agent. The use of a dataset to train an agent is a novel idea in the world of reinforcement learning because, generally, reinforcement learning trains an agent to complete a singular task like the …

The Pawn System: How Procedurally Adaptive Webbed Narratives Create Stories, Steven T. Bordelon

University of New Orleans Theses and Dissertations

This thesis describes the design, implementation, and testing of a novel procedural narrative system called the Procedurally Adaptive Webbed Narrative (PAWN) system. PAWN procedurally generates characters and, responding to choices made by the player, produces more responsive characters and relationships involving the player and these narrative agents. Initially, this thesis discusses other interactive narrative types that exist, such as emergent or event-driven narratives, along with their strengths and weaknesses. It then examines each aspect of PAWN, starting with initial actor generation, then moving to the capturing of game events and translating them into logical objects called Occurrences. These Occurrences are …

Comparative Predictive Analysis Of Stock Performance In The Tech Sector, Asaad Sendi

University of New Orleans Theses and Dissertations

This study compares the performance of deep learning models, including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Transformer, in predicting stock prices across five companies (AAPL, CSCO, META, MSFT, and TSLA) from July 2019 to July 2023. Key findings reveal that GRU models generally exhibit the lowest Mean Absolute Error (MAE), indicating higher precision, particularly notable for CSCO with a remarkably low MAE. While LSTM models often show slightly higher MAE values, they outperform Transformer models in capturing broader trends and variance in stock prices, as evidenced by higher R-squared (R2) values. Transformer models generally exhibit higher MAE …

Choreographing The Rhythms Of Observation: Dynamics For Ranged Observer Bipartite-Unipartite Spatiotemporal (Robust) Networks, Edward A. Holmberg Iv

University of New Orleans Theses and Dissertations

Existing network analysis methods struggle to optimize observer placements in dynamic environments with limited visibility. This dissertation introduces the novel ROBUST (Ranged Observer Bipartite-Unipartite SpatioTemporal) framework, offering a significant advancement in modeling, analyzing, and optimizing observer networks within complex spatiotemporal domains. ROBUST leverages a unique bipartite-unipartite approach, distinguishing between observer and observable entities while incorporating spatial constraints and temporal dynamics.

This research extends spatiotemporal network theory by introducing novel graph-based measures, including myopic degree, spatial closeness centrality, and edge length proportion. These measures, coupled with advanced clustering techniques like Proximal Recurrence, provide insights into network structure, resilience, and the effectiveness …

Analysis Of Forensic Artifacts In Database Memory Using Support Vector Machine, Mahfuzul I. Nissan

University of New Orleans Theses and Dissertations

Memory analysis allows forensic investigators to establish a more complete timeline of system activity using a snapshot of main memory (i.e., RAM). Investigators may rely on such analysis to detect malicious activity and understand the scope of what data was exfiltrated. This is of particular interest in the presence of incomplete or untrusted logs, where a privileged user (or an attacker with such capabilities) can altogether bypass or disable logging. In such instances, a forensic investigator can still rely on the fact that data must ultimately be processed in memory, regardless of the information that is recorded in audit logs. …

Denoising And Deconvolving Sperm Whale Data In The Northern Gulf Of Mexico Using Fourier And Wavelet Techniques, Kendal Mccain Leftwich

University of New Orleans Theses and Dissertations

The use of underwater acoustics can be an important component in obtaining information from the oceans of the world. It is desirable (but difficult) to compile an acoustic catalog of sounds emitted by various underwater objects to complement optical catalogs. For example, the current visual catalog for whale tail flukes of large marine mammals (whales) can identify even individual whales from their individual fluke characteristics. However, since sperm whales, Physeter microcephalus, do not fluke up when they dive, they cannot be identified in this manner. A corresponding acoustic catalog for sperm whale clicks could be compiled to identify individual …

Protein-Protein Interaction Prediction From Language Of Biological Coding, Nayan Howladar

University of New Orleans Theses and Dissertations

Protein-protein interactions in a cell are essential to the characterization and performance of various fundamental biological processes. Due to the tedious, resource-expensive, and time-consuming experimental processes, computational techniques to solve protein pair interaction difficulties have emerged as an active research area in bioinformatics. This research seeks to develop an innovative machine learning-based technique that predicts the interaction of a protein pair based on carefully selected input features and exploits information-rich evolutionary information. We developed a protein-protein interaction predictor, PPILS, that leverages the evolutionary knowledge from the protein language model. We examined several distinct neural network architectures: CNN+LSTM, Transformer, Encoder-Decoder, and …

Ocean Wave Prediction And Characterization For Intelligent Maritime Transportation, Pujan Pokhrel

University of New Orleans Theses and Dissertations

The national Earth System Prediction (ESPC) initiative aims to develop the predictions
for the next generation predictions of atmosphere, ocean, and sea-ice interactions in the scale of days to decades. This dissertation seeks to demonstrate the methods we can use to improve the ESPC models, especially the ocean prediction model. In the application side of the weather forecasts, this dissertation explores imitation learning with constraints to solve combinatorial optimization problems, focusing on the weather routing of surface vessels. Prediction of ocean waves is essential for various purposes, including vessel routing, ocean energy harvesting, agriculture, etc. Since the machine learning approaches …

Parallel Algorithms For Scalable Graph Mining: Applications On Big Data And Machine Learning, Naw Safrin Sattar

University of New Orleans Theses and Dissertations

Parallel computing plays a crucial role in processing large-scale graph data. Complex network analysis is an exciting area of research for many applications in different scientific domains e.g., sociology, biology, online media, recommendation systems and many more. Graph mining is an area of interest with diverse problems from different domains of our daily life. Due to the advancement of data and computing technologies, graph data is growing at an enormous rate, for example, the number of links in social networks is growing every millisecond. Machine/Deep learning plays a significant role for technological accomplishments to work with big data in modern …

Levee Seepage Identification From Aerial Images Using Machine Learning, Sofiane Benkara

University of New Orleans Theses and Dissertations

Levees protect from natural disasters that can threaten human health, infrastructure, and biological systems by protecting low-lying lands near or below sea level from flooding. However, seepage in those levees undermines their structural integrity, leading to failures. Today the United States has approximately over a hundred thousand miles of levee, many of which are reaching or have surpassed their initial design life. Given the concern, there is a need to develop reliable, rapid, and non-intrusive levee monitoring systems to detect the presence of seepage. This study explores the use of Deep Convolutional Neural Network (DCNN) integrated with Discrete Cosine Transform …

Video Games, Grief, And The Character Link System, Nam Nguyen

University of New Orleans Theses and Dissertations

Grief can encompass more than just the loss of real-life people. It can be felt with the loss of a pet, changes in daily structure, and even the loss of video game characters. The topic of grief related to video games and video game characters comes at a time when games as a service (GaaS) continue to increase in popularity and the phenomenon where these games also inevitably terminate service. To combat this unique form of grief, the Character LINK System was created as a tool that uses simple natural language processing (NLP) techniques to offer support to the bereaved …

Analysis Of Residual Neural Networks For Marine Mammal Classification Using Multi-Channel Spectrograms, Daniel T. Murphy

University of New Orleans Theses and Dissertations

Surveys of marine mammal populations are an essential part of monitoring the welfare of these animals and their ecosystems. Marine mammal vocalizations provide a reliable method of identifying most species, but passive acoustic monitoring of underwater audio may generate large quantities of data that exceed the capacity of human classifiers. Preprocessing and machine learning techniques provide a method of automating the classification process. In this study, we explore machine learning approaches to vocalization classification using convolutional neural networks with residual learning. Optimal parameters for noise-removal, spectrographic window functions, preprocessing augmentations, and multi-channel spectrogram generation are derived through a series of …

Machine Learning For Terminal Procedure Chart Change Detection, Anthony M. Marchiafava

University of New Orleans Theses and Dissertations

Terminal Procedure Charts are a constantly updated and necessary tool for aircraft personnel to approach and take off from airport runways safely. Detecting changes within these charts is a time-consuming and laborious process. Here machine learning techniques were used to predict regions of change in charts based on detecting the charts image regions and comparing features extracted from those regions. Outlined are methodologies to detect differences between two separate charts to produce images with changed regions clearly indicated. Both more conventional computer vision and machine learning techniques were applied. For images with minor shifts, the proposed model is able to …

Convolutional Neural Networks For Deflate Data Encoding Classification Of High Entropy File Fragments, Nehal Ameen

University of New Orleans Theses and Dissertations

Data reconstruction is significantly improved in terms of speed and accuracy by reliable data encoding fragment classification. To date, work on this problem has been successful with file structures of low entropy that contain sparse data, such as large tables or logs. Classifying compressed, encrypted, and random data that exhibit high entropy is an inherently difficult problem that requires more advanced classification approaches. We explore the ability of convolutional neural networks and word embeddings to classify deflate data encoding of high entropy file fragments after establishing ground truth using controlled datasets. Our model is designed to either successfully classify file …

The Kati Module System: Modular Design For Delivering Character Focused Dialogue In Games, Stephen J. Marcel

University of New Orleans Theses and Dissertations

The Kati Module System is an interconnected set of programming modules intended to facilitate dynamic text authoring for interactive experiences (for example, games). It is a long-standing goal for interactive experiences to dynamically adapt their textual output based on the user or player's choices and predilections, but to account for this vast possibility space requires an amount of authoring that is frequently untenable, especially for small studios. Advances in machine learning have produced incredible progress in the field of Natural Language Generation (NLG). Though this produces impressive surface level text, it does so without an internal representation that can be …

Sounds Of Silence: A Study Of Stability And Diversity Of Web Audio Fingerprints, Shekhar Chalise

University of New Orleans Theses and Dissertations

Browser fingerprinting presents a grave threat to privacy as it allows user tracking even in private browsing modes. Prior measurement studies on HTML5-based fingerprinting have been limited to Canvas and WebGL but not Web Audio APIs. We aim to fill this gap by conducting the first large-scale systematic study of web audio fingerprints and studying their stability as well as diversity properties. Using MTurk and social media platforms, we collected 8 different audio fingerprints from 694 users.

Firstly, we show that the audio fingerprints are unstable unlike other fingerprinting methods with some users having as many as 20 different fingerprints. …

Machine Learning Model Selection For Predicting Global Bathymetry, Nicholas P. Moran

University of New Orleans Theses and Dissertations

This work is concerned with the viability of Machine Learning (ML) in training models for predicting global bathymetry, and whether there is a best fit model for predicting that bathymetry. The desired result is an investigation of the ability for ML to be used in future prediction models and to experiment with multiple trained models to determine an optimum selection. Ocean features were aggregated from a set of external studies and placed into two minute spatial grids representing the earth's oceans. A set of regression models, classification models, and a novel classification model were then fit to this data and …

Using High-Performance Computing Profilers To Understand The Performance Of Graph Algorithms, Costain Nachuma

University of New Orleans Theses and Dissertations

An algorithm designer working with parallel computing systems should know how the characteristics of their implemented algorithm affects various performance aspects of their parallel program. It would be beneficial to these designers if each algorithm came with a specific set of standards that identified which algorithms worked better for a specified system. Therefore, the goal of this paper is to take implementations of four graphing algorithms, extract their features such as memory consumption, scalability using profilers (Vtunes /Tau) to determine which algorithms work to their fullest potential in one of the three systems: GPU, shared memory system, or distributed memory …

Detecting Convergence Zone Paths In Acoustic Model Outputs Using Machine Learning, Michael Sinegar

University of New Orleans Theses and Dissertations

Sound propagated underwater can possibly travel according to several different patterns. One such pattern, convergence zone (CZ), is the main focus of this thesis. This thesis presents an ArcGIS-based tool to easily choose specific points in the Atlantic Ocean based on latitude and longitude, then gather data about the propagation of sound at that point. In addition to this, a mini-app that generates machine learning datasets was created. It easily allows for one to label thousands of images in a short amount of time. A thousand CZ and a thousand non-CZ images were used to train a machine learning algorithm …

Accelerating The Information-Theoretic Approach Of Community Detection Using Distributed And Hybrid Memory Parallel Schemes, Md Abdul Motaleb Faysal

University of New Orleans Theses and Dissertations

There are several approaches for discovering communities in a network (graph). Despite being approximating in nature, discovering communities based on the laws of Information Theory has a proven standard of accuracy. The information-theoretic algorithm known as Infomap developed a decade ago for detecting communities, did not foresee the tremendous growth of social networking, multimedia, and massive information boom. To discover communities in massive networks, we have designed a distributed-memory-parallel Infomap in the MPI framework. Our design reaches scalability of over 500 processes capable of processing networks with millions of edges while maintaining quality comparable to the sequential Infomap. We have …

Ship Detection Feature Analysis In Optical Satellite Imagery Through Machine Learning Applications, Sylvia Charchut

University of New Orleans Theses and Dissertations

Ship detection remains an important challenge within the government and the commercial industry. Current research has focused on deep learning and has found high success with large labeled datasets. However, deep learning becomes insufficient for limited datasets as well as when explainability is required. There exist scenarios in which explainability and human-in-the-loop processing are needed, such as in naval applications. In these scenarios, handcrafted features and traditional classification algorithms can be useful. This research aims at analyzing multiple textures and statistical features on a small optical satellite imagery dataset. The feature analysis consists of Haar-like features, Haralick features, Hu moments, …

A Domain Specific Language For Digital Forensics And Incident Response Analysis, Christopher D. Stelly

University of New Orleans Theses and Dissertations

One of the longstanding conceptual problems in digital forensics is the dichotomy between the need for verifiable and reproducible forensic investigations, and the lack of practical mechanisms to accomplish them. With nearly four decades of professional digital forensic practice, investigator notes are still the primary source of reproducibility information, and much of it is tied to the functions of specific, often proprietary, tools.

The lack of a formal means of specification for digital forensic operations results in three major problems. Specifically, there is a critical lack of:

a) standardized and automated means to scientifically verify accuracy of digital forensic tools; …

The Effects Of Automated Grading On Computer Science Courses At The University Of New Orleans, Jerod F A Dunbar

University of New Orleans Theses and Dissertations

This is a study of the impacts of the incorporation, into certain points of the Computer Science degree program at the University of New Orleans, of Course Management software with an Autograding component. The software in question, developed at Carnegie Mellon University, is called “Autolab.” We begin by dissecting Autolab in order to gain an understanding of its inner workings. We can then take out understanding of its functionality and apply that to an examination of fundamental changes to courses in the time since they incorporated the software. With that, we then compare Drop, Failure, Withdrawal rate data from before …

Multi-Agent Narrative Experience Management As Story Graph Pruning, Edward T. Garcia

University of New Orleans Theses and Dissertations

In this thesis I describe a method where an experience manager chooses actions for non-player characters (NPCs) in intelligent interactive narratives through story graph representation and pruning. The space of all stories can be represented as a story graph where nodes are states and edges are actions. By shaping the domain as a story graph, experience manager decisions can be made by pruning edges. Starting with a full graph, I apply a set of pruning strategies that will allow the narrative to be finishable, NPCs to act believably, and the player to be responsible for how the story unfolds. By …

A Machine Learning Assessment To Predict The Sediment Transport Rate Under Oscillating Sheet Flow Conditions, Huy Vu

Senior Honors Theses

The two-phase flow approach has been the conventional method designed to study the sediment transport rate. Due to the complexity of sediment transport, the precisely numerical models computed from that approach require initial assumptions and, as a result, may not yield accurate output for all conditions. This research work proposes that Machine Learning algorithms can be an alternative way to predict the processes of sediment transport in two-dimensional directions under oscillating sheet flow conditions, by utilizing the available dataset of the SedFoam multidimensional two-phase model. The assessment utilized linear regression and gradient boosting algorithm to analyze the lowest average mean …

Prediction Of Hierarchical Classification Of Transposable Elements Using Machine Learning Techniques, Manisha Panta

University of New Orleans Theses and Dissertations

Transposable Elements (TEs) or jumping genes are the DNA sequences that have an intrinsic capability to move within a host genome from one genomic location to another. Studies show that the presence of a TE within or adjacent to a functional gene may alter its expression. TEs can also cause an increase in the rate of mutation and can even promote gross genetic arrangements. Thus, the proper classification of the identified jumping genes is important to understand their genetic and evolutionary effects. While computational methods have been developed that perform either binary classification or multi-label classification of TEs, few studies …

Effective Statistical Energy Function Based Protein Un/Structure Prediction, Avdesh Mishra

University of New Orleans Theses and Dissertations

Proteins are an important component of living organisms, composed of one or more polypeptide chains, each containing hundreds or even thousands of amino acids of 20 standard types. The structure of a protein from the sequence determines crucial functions of proteins such as initiating metabolic reactions, DNA replication, cell signaling, and transporting molecules. In the past, proteins were considered to always have a well-defined stable shape (structured proteins), however, it has recently been shown that there exist intrinsically disordered proteins (IDPs), which lack a fixed or ordered 3D structure, have dynamic characteristics and therefore, exist in multiple states. Based on …

Scalable Community Detection Using Distributed Louvain Algorithm, Naw Safrin Sattar

University of New Orleans Theses and Dissertations

Community detection (or clustering) in large-scale graph is an important problem in graph mining. Communities reveal interesting characteristics of a network. Louvain is an efficient sequential algorithm but fails to scale emerging large-scale data. Developing distributed-memory parallel algorithms is challenging because of inter-process communication and load-balancing issues. In this work, we design a shared memory-based algorithm using OpenMP, which shows a 4-fold speedup but is limited to available physical cores. Our second algorithm is an MPI-based parallel algorithm that scales to a moderate number of processors. We also implement a hybrid algorithm combining both. Finally, we incorporate dynamic load-balancing in …

Detection Of Sand Boils From Images Using Machine Learning Approaches, Aditi S. Kuchi

University of New Orleans Theses and Dissertations

Levees provide protection for vast amounts of commercial and residential properties. However, these structures degrade over time, due to the impact of severe weather, sand boils, subsidence of land, seepage, etc. In this research, we focus on detecting sand boils. Sand boils occur when water under pressure wells up to the surface through a bed of sand. These make levees especially vulnerable. Object detection is a good approach to confirm the presence of sand boils from satellite or drone imagery, which can be utilized to assist in the automated levee monitoring methodology. Since sand boils have distinct features, applying object …

Measuring Interactive Narrative Quality With Experience Management As Story Graph Pruning, Jean-Paul Jeunesse

Senior Honors Theses

An interactive narrative in a virtual environment is created through player and system interaction, often through an experience manager controlling the actions of all non-player characters (NPCs). Thus, the narrative (and its quality) is entirely dependent on a conflicting combination of unpredictability from the player and a controlled environment that must react to this unpredictability. Ideally, the experience manager should decide NPC actions in a way that never limits player freedom and shows the NPCs acting in believable manners to create a story that can be meaningfully affected by the player and feels organic. One solution to this is to …