Physical Sciences and Mathematics Commons^™

Feature Selection And Analysis For Standard Machine Learning Classification Of Audio Beehive Samples, Chelsi Gupta

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

The beekeepers need to inspect their hives regularly in order to protect them from various stressors. Manual inspection of hives require a lot of time and effort. Hence, many researchers have started using electronic beehive monitoring (EBM) systems to collect critical information from beehives, so as to alert the beekeepers of possible threats to the hive. EBM collects information by applying multiple sensors into the hive. The sensors collect information in the form of video, audio or temperature data from the hives.

This thesis involves the automatic classification of audio samples from a beehive into bee buzzing, cricket chirping and …

Enhancing Scalability In Genetic Programming With Adaptable Constraints, Type Constraints And Automatically Defined Functions, George Gerules

Dissertations

Genetic Programming is a type of biological inspired machine learning. It is composed of a population of stochastic individuals. Those individuals can exchange portions of themselves with others in the population through the crossover operation that draws its inspiration from biology. Other biologically inspired operations include mutation and reproduction. The form an individual takes can be many things. It, however, is represented most of the time as a computer program. Constructing correct efficient programs can be notoriously difficult. Various grammar, typing, function constraint, or counting mechanisms can guide creation and evolution of those individuals. These mechanisms can reduce search space …

Df 2.0: An Automated, Privacy Preserving, And Efficient Digital Forensic Framework That Leverages Machine Learning For Evidence Prediction And Privacy Evaluation, Robin Verma, Jayaprakash Govindaraj Dr, Saheb Chhabra, Gaurav Gupta

Journal of Digital Forensics, Security and Law

The current state of digital forensic investigation is continuously challenged by the rapid technological changes, the increase in the use of digital devices (both the heterogeneity and the count), and the sheer volume of data that these devices could contain. Although data privacy protection is not a performance measure, however, preventing privacy violations during the digital forensic investigation, is also a big challenge. With a perception that the completeness of investigation and the data privacy preservation are incompatible with each other, the researchers have provided solutions to address the above-stated challenges that either focus on the effectiveness of the investigation …

Supervised Machine Learning Models For Fake News Detection, Andrea Lopez, Adelo Vieira, Zafar Ahsan, Farooq Sabib, Shirley Marinho

ICT

Fake news or the distribution of disinformation has become one of the most challenging issues in society. News and information are churned out across online websites and platforms in real-time, with little or no way for the viewing public to determine what is real or manufactured. But an awareness of what we are consuming online is becoming apparent and efforts are underway to explore how we separate fake content from genuine and truthful information. The most challenging part of fake news is determining how to spot it. In technology, there are ways to help us do this. Supervised machine learning …

Multi-Resolution Models For Learning Multilevel Abstract Representation With Application To Information Retrieval, Tolgahan Cakaloglu

Theses and Dissertations

Deep language models learning a hierarchical representation proved to be a powerful tool for natural language processing, text mining, and information retrieval tasks. However, more specifically, representations that perform well for ad-hoc retrieval must capture semantic meaning at different levels of abstraction or context-scopes. The primary goal of ad-hoc retrieval is to find relevant documents satisfying the information need posted in a natural language query. It requires a good understanding of the query and all the documents in a corpus, which is difficult because the meaning of natural language texts depends on the context, syntax, and semantics. In this dissertation, …

Identifying Hourly Traffic Patterns With Python Deep Learning, Christopher L. Leavitt

Computer Engineering

This project was designed to explore and analyze the potential abilities and usefulness of applying machine learning models to data collected by parking sensors at a major metro shopping mall. By examining patterns in rates at which customer enter and exit parking garages on the campus of the Bellevue Collection shopping mall in Bellevue, Washington, a recurrent neural network will use data points from the previous hours will be trained to forecast future trends.

Classifying Challenging Behaviors In Autism Spectrum Disorder With Neural Document Embeddings, Abigail Atchison

Computational and Data Sciences (MS) Theses

The understanding and treatment of challenging behaviors in individuals with Autism Spectrum Disorder is paramount to enabling the success of behavioral therapy; an essential step in this process being the labeling of challenging behaviors demonstrated in therapy sessions. These manifestations differ across individuals and within individuals over time and thus, the appropriate classification of a challenging behavior when considering purely qualitative factors can be unclear. In this thesis we seek to add quantitative depth to this otherwise qualitative task of challenging behavior classification. We do so through the application of natural language processing techniques to behavioral descriptions extracted from the …

Using Satellite-Based Hydro-Climate Variables And Machine Learning For Streamflow Modeling At Various Scales In The Upper Mississippi River Basin, Dongjae Kwon

Theses and Dissertations

Streamflow data are essential to study the hydrologic cycle and to attain appropriate water resource management policies. However, the availability of gauge data is limited due to various reasons such as economic, political, instrumental malfunctioning, and poor spatial distribution. Although streamflow can be simulated by process-based and machine learning approaches, applicability is limited due to intensive modeling effort, or its black-box nature, respectively. Here, we introduce a machine learning (Boosted Regression Tree (BRT)) approach based on remote sensing data to simulate monthly streamflow for three of varying sizes watersheds in the Upper Mississippi River Basin (UMRB). By integrating spatial land …

Emulation Vs Instrumentation For Android Malware Detection, Anukriti Sinha

Master's Projects

In resource constrained devices, malware detection is typically based on offline analysis using emulation. In previous work it has been claimed that such emulation fails for a significant percentage of Android malware because well-designed malware detects that the code is being emulated. An alternative to emulation is malware analysis based on code that is executing on an actual Android device. In this research, we collect features from a corpus of Android malware using both emulation and on-phone instrumentation. We train machine learning models based on emulated features and also train models based on features collected via instrumentation, and we compare …

Classifying Classic Ciphers Using Machine Learning, Nivedhitha Ramarathnam Krishna

Master's Projects

We consider the problem of identifying the classic cipher that was used to generate a given ciphertext message. We assume that the plaintext is English and we restrict our attention to ciphertext consisting only of alphabetic characters. Among the classic ciphers considered are the simple substitution, Vigenère cipher, playfair cipher, and column transposition cipher. The problem of classification is approached in two ways. The first method uses support vector machines (SVM) trained directly on ciphertext to classify the ciphers. In the second approach, we train hidden Markov models (HMM) on each ciphertext message, then use these trained HMMs as features …

Differential Estimation Of Audiograms Using Gaussian Process Active Model Selection, Trevor Larsen

McKelvey School of Engineering Graduate Student Theses & Dissertations

Classical methods for psychometric function estimation either require excessive resources to perform, as in the method of constants, or produce only a low resolution approximation of the target psychometric function, as in adaptive staircase or up-down procedures. This thesis makes two primary contributions to the estimation of the audiogram, a clinically relevant psychometric function estimated by querying a patient’s for audibility of a collection of tones. First, it covers the implementation of a Gaussian process model for learning an audiogram using another audiogram as a prior belief to speed up the learning procedure. Second, it implements a use case of …

Bias Reduction In Machine Learning Classifiers For Spatiotemporal Analysis Of Coral Reefs Using Remote Sensing Images, Justin J. Gapper

Computational and Data Sciences (PhD) Dissertations

This dissertation is an evaluation of the generalization characteristics of machine learning classifiers as applied to the detection of coral reefs using remote sensing images. Three scientific studies have been conducted as part of this research: 1) Evaluation of Spatial Generalization Characteristics of a Robust Classifier as Applied to Coral Reef Habitats in Remote Islands of the Pacific Ocean 2) Coral Reef Change Detection in Remote Pacific Islands using Support Vector Machine Classifiers 3) A Generalized Machine Learning Classifier for Spatiotemporal Analysis of Coral Reefs in the Red Sea. The aim of this dissertation is to propose and evaluate a …

Repairing Landsat Satellite Imagery Using Deep Machine Learning Techniques, Griffin J. Lane, Patricia Goresen, Robert Slater

SMU Data Science Review

Satellite Imagery is one of the most widely used sources to analyze geographic features and environments in the world. The data gathered from satellites are used to quantify many vital problems facing our society, such as the impact of natural disasters, shore erosion, rising water levels, and urban growth rates. In this paper, we construct machine learning and deep learning algorithms for repairing anomalies in the Landsat satellite imagery data which arise for various reasons ranging from cloud obstruction to satellite malfunctions. The accuracy of GIS data is crucial to ensuring the models produced from such data are as close …

Visualization And Machine Learning Techniques For Nasa’S Em-1 Big Data Problem, Antonio P. Garza Iii, Jose Quinonez, Misael Santana, Nibhrat Lohia

SMU Data Science Review

In this paper, we help NASA solve three Exploration Mission-1 (EM-1) challenges: data storage, computation time, and visualization of complex data. NASA is studying one year of trajectory data to determine available launch opportunities (about 90TBs of data). We improve data storage by introducing a cloud-based solution that provides elasticity and server upgrades. This migration will save $120k in infrastructure costs every four years, and potentially avoid schedule slips. Additionally, it increases computational efficiency by 125%. We further enhance computation via machine learning techniques that use the classic orbital elements to predict valid trajectories. Our machine learning model decreases trajectory …

Machine Learning Pipeline For Exoplanet Classification, George Clayton Sturrock, Brychan Manry, Sohail Rafiqi

SMU Data Science Review

Planet identification has typically been a tasked performed exclusively by teams of astronomers and astrophysicists using methods and tools accessible only to those with years of academic education and training. NASA’s Exoplanet Exploration program has introduced modern satellites capable of capturing a vast array of data regarding celestial objects of interest to assist with researching these objects. The availability of satellite data has opened up the task of planet identification to individuals capable of writing and interpreting machine learning models. In this study, several classification models and datasets are utilized to assign a probability of an observation being an exoplanet. …

Commonsense Knowledge In Sentiment Analysis Of Ordinance Reactions For Smart Governance, Manish Puri

Theses, Dissertations and Culminating Projects

Smart Governance is an emerging research area which has attracted scientific as well as policy interests, and aims to improve collaboration between government and citizens, as well as other stakeholders. Our project aims to enable lawmakers to incorporate data driven decision making in enacting ordinances. Our first objective is to create a mechanism for mapping ordinances (local laws) and tweets to Smart City Characteristics (SCC). The use of SCC has allowed us to create a mapping between a huge number of ordinances and tweets, and the use of Commonsense Knowledge (CSK) has allowed us to utilize human judgment in mapping. …

Supervised Machine Learning Models For Fake News Detection, Gofaas Group, Andrea Lopez, Adelo Vieira, Zafar Ahsan, Farooq Saqib, Shirley Marinho

ICT

Fake news or the distribution of disinformation has become one of the most challenging issues in society. News and information are churned out across online websites and platforms in real-time, with little or no way for the viewing public to determine what is real or manufactured. But an awareness of what we are consuming online is becoming apparent and efforts are underway to explore how we separate fake content from genuine and truthful information.

The most challenging part of fake news is determining how to spot it. In technology, there are ways to help us do this. Supervised machine learning …

Deep Embedding Kernel, Linh Le

Doctor of Data Science and Analytics Dissertations

Kernel methods and deep learning are two major branches of machine learning that have achieved numerous successes in both analytics and artificial intelligence. While having their own unique characteristics, both branches work through mapping data to a feature space that is supposedly more favorable towards the given task. This dissertation addresses the strengths and weaknesses of each mapping method through combining them and forming a family of novel deep architectures that center around the Deep Embedding Kernel (DEK). In short, DEK is a realization of a kernel function through a newly deep architecture. The mapping in DEK is both implicit …

Predicting Win Rates In Competitive Overwatchtm, Andrea Sibley

Mathematics Senior Capstone Papers

OverwatchTM is a video game published by Blizzard Entertainment R where two teams comprised of six people each compete against one another to accomplish a specific goal. The goal of each game is dependent on which map is being played. The maps are divided into four categories: Assault, Escort, Control, and Hybrid. A data set comprised of 3000 games of competitive OverwatchTM is used to determine how likely a team is to win their match. The factors used to determine the likelihood of winning are the map type and the skill ranking for each team. The data set is pre-processed …

Neural Machine Translation, Quinn M. Lanners, Thomas Laurent

Honors Thesis

Neural Machine Translation is the primary algorithm used in industry to perform machine translation. This state-of-the-art algorithm is an application of deep learning in which massive datasets of translated sentences are used to train a model capable of translating between any two languages. The architecture behind neural machine translation is composed of two recurrent neural networks used together in tandem to create an Encoder Decoder structure. Attention mechanisms have recently been developed to further increase the accuracy of these models. In this senior thesis, the various parts of Neural Machine Translation are explored towards the eventual creation of a tutorial …

A Study Of Face Embedding In Face Recognition, Khanh Duc Le

Master's Theses

Face Recognition has been a long-standing topic in computer vision and pattern recognition field because of its wide and important applications in our daily lives such as surveillance system, access control, and so on. The current modern face recognition model, which keeps only a couple of images per person in the database, can now recognize a face with high accuracy. Moreover, the model does not need to be retrained every time a new person is added to the database.

By using the face dataset from Digital Democracy, the thesis will explore the capability of this model by comparing it with …

Cs04all: Machine Learning Module, Hunter R. Johnson

Open Educational Resources

These are materials that may be used in a CS0 course as a light introduction to machine learning.

The materials are mostly Jupyter notebooks which contain a combination of labwork and lecture notes. There are notebooks on Classification, An Introduction to Numpy, and An Introduction to Pandas.

There are also two assessments that could be assigned to students. One is an essay assignment in which students are asked to read and respond to an article on machine bias. The other is a lab-like exercise in which students use pandas and numpy to extract useful information about subway ridership in NYC. …

Dish: Democracy In State Houses, Nicholas A. Russo

Master's Theses

In our current political climate, state level legislators have become increasingly impor- tant. Due to cuts in funding and growing focus at the national level, public oversight for these legislators has drastically decreased. This makes it difficult for citizens and activists to understand the relationships and commonalities between legislators. This thesis provides three contributions to address this issue. First, we created a data set containing over 1200 features focused on a legislator’s activity on bills. Second, we created embeddings that represented a legislator’s level of activity and engagement for a given bill using a custom model called Democracy2Vec. Third, we …

An Evaluation Of Training Size Impact On Validation Accuracy For Optimized Convolutional Neural Networks, Jostein Barry-Straume, Adam Tschannen, Daniel W. Engels, Edward Fine

SMU Data Science Review

In this paper, we present an evaluation of training size impact on validation accuracy for an optimized Convolutional Neural Network (CNN). CNNs are currently the state-of-the-art architecture for object classification tasks. We used Amazon’s machine learning ecosystem to train and test 648 models to find the optimal hyperparameters with which to apply a CNN towards the Fashion-MNIST (Mixed National Institute of Standards and Technology) dataset. We were able to realize a validation accuracy of 90% by using only 40% of the original data. We found that hidden layers appear to have had zero impact on validation accuracy, whereas the neural …

Comparisons Of Performance Between Quantum And Classical Machine Learning, Christopher Havenstein, Damarcus Thomas, Swami Chandrasekaran

SMU Data Science Review

In this paper, we present a performance comparison of machine learning algorithms executed on traditional and quantum computers. Quantum computing has potential of achieving incredible results for certain types of problems, and we explore if it can be applied to machine learning. First, we identified quantum machine learning algorithms with reproducible code and had classical machine learning counterparts. Then, we found relevant data sets with which we tested the comparable quantum and classical machine learning algorithm's performance. We evaluated performance with algorithm execution time and accuracy. We found that quantum variational support vector machines in some cases had higher accuracy …

Comparative Study Of Sentiment Analysis With Product Reviews Using Machine Learning And Lexicon-Based Approaches, Heidi Nguyen, Aravind Veluchamy, Mamadou Diop, Rashed Iqbal

SMU Data Science Review

In this paper, we present a comparative study of text sentiment classification models using term frequency inverse document frequency vectorization in both supervised machine learning and lexicon-based techniques. There have been multiple promising machine learning and lexicon-based techniques, but the relative goodness of each approach on specific types of problems is not well understood. In order to offer researchers comprehensive insights, we compare a total of six algorithms to each other. The three machine learning algorithms are: Logistic Regression (LR), Support Vector Machine (SVM), and Gradient Boosting. The three lexicon-based algorithms are: Valence Aware Dictionary and Sentiment Reasoner (VADER), Pattern, …

Improving Vix Futures Forecasts Using Machine Learning Methods, James Hosker, Slobodan Djurdjevic, Hieu Nguyen, Robert Slater

SMU Data Science Review

The problem of forecasting market volatility is a difficult task for most fund managers. Volatility forecasts are used for risk management, alpha (risk) trading, and the reduction of trading friction. Improving the forecasts of future market volatility assists fund managers in adding or reducing risk in their portfolios as well as in increasing hedges to protect their portfolios in anticipation of a market sell-off event. Our analysis compares three existing financial models that forecast future market volatility using the Chicago Board Options Exchange Volatility Index (VIX) to six machine/deep learning supervised regression methods. This analysis determines which models provide best …

A Machine Learning Recommender Model For Ride Sharing Based On Rider Characteristics And User Threshold Time, Govind Pramod Yatnalkar

Theses, Dissertations and Capstones

In the present age, human life is prospering incredibly due to the 4^th Industrial Revolution or The Age of Digitization and Computing. The ubiquitous availability of the Internet and advanced computing systems have resulted in the rapid development of smart cities. From connected devices to live vehicle tracking, technology is taking the field of transportation to a new level. An essential part of the transportation domain in smart cities is Ride Sharing. It is an excellent solution to issues like pollution, traffic, and the rapid consumption of fuel. Even though Ride Sharing has several benefits, the current usage is …

Abso2luteu-Net: Tissue Oxygenation Calculation Using Photoacoustic Imaging And Convolutional Neural Networks, Kevin Hoffer-Hawlik, Geoffrey P. Luke

ENGS 88 Honors Thesis (AB Students)

Photoacoustic (PA) imaging uses incident light to generate ultrasound signals within tissues. Using PA imaging to accurately measure hemoglobin concentration and calculate oxygenation (sO₂) requires prior tissue knowledge and costly computational methods. However, this thesis shows that machine learning algorithms can accurately and quickly estimate sO₂. absO₂luteU-Net, a convolutional neural network, was trained on Monte Carlo simulated multispectral PA data and predicted sO₂ with higher accuracy compared to simple linear unmixing, suggesting machine learning can solve the fluence estimation problem. This project was funded by the Kaminsky Family Fund and the Neukom Institute.

Data Driven Approach To Characterize And Forecast The Impact Of Freeway Work Zones On Mobility Using Probe Vehicle Data, Mohsen Kamyab

Wayne State University Dissertations

The presence of work zones on freeways causes traffic congestion and creates hazardous conditions for commuters and construction workers. Traffic congestion resulting from work zones causes negative impacts on traffic mobility (delay), the environment (vehicle emissions), and safety when stopped or slowed vehicles become vulnerable to rear-end collisions. Addressing these concerns, a data-driven approach was utilized to develop methodologies to measure, predict, and characterize the impact work zones have on Michigan interstates. This study used probe vehicle data, collected from GPS devices in vehicles, as the primary source for mobility data. This data was used to fulfill three objectives: develop …