Matching Items (756)
Filtering by
- Creators: Computer Science and Engineering Program
DescriptionDesigning and implementing components for a new SMS messaging strategy for Go Together Inc. Includes ensuring compliance with federal messaging regulations, a schoolwide notifications/announcements console, feedback reception for completed trips via automated messaging, and user segmentation and automated messaging.
ContributorsJaramillo, Evan (Author) / Baumann, Alicia (Thesis director) / Moore, Kimberly (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05
Description
This applied thesis project presents an integrated Financial Market Prediction
System that provides cross-asset analysis through advanced machine learning techniques.
By seamlessly unifying predictive analytics across stocks, cryptocurrencies, forex,
commodities, and real estate markets, the system transcends traditional siloed approaches
to financial forecasting. Leveraging machine learning ensemble methods including
Gradient Boosting, Linear Regression and Random Forest algorithms. The architecture
incorporates a sophisticated constellation of over 25 technical indicators, asset-specific
feature engineering, and a recursive prediction methodology that dynamically calibrates
confidence intervals based on temporal distance and market volatility.
The system's elegant backtesting framework employs walk-forward validation
across multiple market regimes, delivering robust performance metrics while effectively
mitigating overfitting risks inherent in financial modeling. Through an intuitive yet
powerful interactive interface, users can effortlessly navigate complex market dynamics,
visualize probabilistic forecasts, and gain unprecedented insights into feature importance
hierarchies that drive market movements. Beyond mere prediction, the platform serves as
an educational cornerstone, seamlessly integrating a comprehensive financial knowledge
base with real-time analytical capabilities.
ContributorsTewary, Ganap Ashit (Author) / Menees, Jodi (Thesis director) / Srinivasan, Aravind (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05
Description
P.O.F.T.S. is an open-source fitness-tracking platform designed with transparent data privacy, a remarkable level of customization, and frequent user interaction in mind. Unlike other commercial fitness apps on the current market, P.O.F.T.S. gives the users total control over their data for a personalized fitness experience, including workout recommendations, social leaderboards, and devices synced with wearables.
P.O.F.T.S. (Personalized Open-Source Fitness Tracking Software)
ContributorsAlfred Vijay, Heinric (Author) / Mejari, Vikas (Co-author) / Byrne, Jared (Thesis director) / Howell, Travis (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor) / School of Mathematical and Statistical Sciences (Contributor)
Created2025-05
Description
This thesis introduces an intelligent database system that harnesses Natural Language Processing (NLP) and Machine Learning (ML) to enable seamless querying and visualization of sports data. Centered on the English Premier League—the top tier of English football—the system empowers users to interact with complex datasets through simple natural language queries. These queries are automatically translated into structured SQL commands, eliminating the need for technical expertise and making data retrieval more accessible.
In addition to flexible querying, the system supports dynamic data visualization, presenting results in user-specified formats such as tables, charts, or graphs. By integrating NLP and ML, the system streamlines the end-to-end process of data access, analysis, and presentation. This not only enhances the usability of sports data for analysts, researchers, and enthusiasts but also promotes data-driven exploration and insight generation. The proposed system represents a step toward democratizing sports analytics by bridging the gap between natural language understanding and structured data querying, enabling richer, more intuitive interactions with complex information.
ContributorsMartinez, Sebastian (Author) / Gupta, Vivek (Thesis director) / Bryan, Chris (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05
Description
As a brand-new social media platform, Ink-bloom has analytics requirements that fall outside of the normal range covered by existing analytics tools. Additionally, the lack of analytics and data science expertise on the permanent Ink-bloom team necessitates a simpler approach to analytics presentation. As such, this thesis aims to produce an easy to use, integrated analytics tool for use in the product testing stage of the Ink-bloom website. It includes features such as unique visit tracking, bounce-rate calculation, and a dynamically generated clustering algorithm to group user behaviors across the platform
ContributorsGoswick, Logan (Author) / Meuth, Ryan (Thesis director) / Hsu, Ray (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05
Description
This thesis presents Translatica, a modular speech-to-speech translation (S2ST) system
that preserves both linguistic meaning and the speaker’s vocal identity across languages. Alongside developing a working prototype, this work surveys the landscape of S2ST methods and motivates the choice of a modular architecture over direct approaches, emphasizing flexibility, interpretability, and voice fidelity. The system combines state-of-the-art tools in transcription, translation, and voice synthesis to enable expressive, speaker-preserving dubbing of prerecorded videos. Through implementation and evaluation, the thesis explores the trade-offs between accuracy, latency, and control, demonstrating how modular design enables customization for diverse use cases. Future work includes real-time translation, enhanced speaker tracking, and applications in education and live media.
ContributorsJhaj, Baaz (Author) / Ramani, Krishna (Co-author) / Hsu, Jeffrey (Co-author) / Osburn, Steven (Thesis director) / Zhu, Haolin (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05
Description
This paper investigates the relationship between professional sports teams and local crime rates in their host cities, attempting to address a gap in the literature that predominantly focuses on the economic impact of sports. Through statistical analysis of daily crime data from three paragons of United States sports culture (Dallas, Boston, and Philadelphia), this research will examine relationships between home games for NBA and NFL teams and local crime rates. Specifically, it will look at data from 2020, and compare the relationship between the presence of home games and the results of those games to local felonies, misdemeanors, and violations. The findings appear to find notable patterns, in that cities consistently demonstrate statistically significant reductions in felony crimes on game days. While score results and margins showed minimal correlation with crime rates, the presence of a home game was associated with the strongest reduction in crime rates. Overall, this paper contributes to understanding the complex and incredibly nuanced relationship between professional sports and crime, indicating that game days may have a temporary reduction in crime.
ContributorsNair, Rohan (Author) / Meuth, Ryan (Thesis director) / Mahzabeen, Sabiha (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor) / Economics Program in CLAS (Contributor) / School of Mathematical and Statistical Sciences (Contributor) / Department of Information Systems (Contributor)
Created2025-05
Description
Artificial intelligence (AI) has rapidly progressed over the past couple of years, sparking worldwide concerns about whether these new technologies will take human jobs, especially in the film industry. This thesis delves deeper into the ethical implications of AI in filmmaking and its potential to change the film industry as we know it. AI technologies have found ways to streamline production processes, creating benefits such as cost efficiency for small-budget filmmakers. Although these developments can optimize pre-production, production, and post-production processes, it causes distress with factors such as ethical use of information, job security, and creative integrity. A qualitative methodology incorporating literature reviews and case studies of visual media, is used to explore the dual-sided sword AI can be: democratizing the filmmaking process while also making people question the authenticity of the new age of cinema. This study explores whether AI can maintain a human element in art without ruining authorship in AI-generated scripts, or the use of deep fakes. Recommendations are made to ensure transparency and consent are always prioritized when innovating, to maintain ethical integrity. This thesis argues that while AI can support filmmakers through creative inspiration, and streamlining certain processes, it should never completely replace human authorship. AI is there to complement, not supersede the authentic storytelling that defines film. By highlighting these obstacles, this thesis can add to the ongoing conversation about whether AI should be involved in cinema's future and our obligation to establish equitable and progressive regulations.
ContributorsPartha, Ajay (Author) / Malpe, Adwith (Thesis director) / Wheatley, Abby (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor) / Dean, W.P. Carey School of Business (Contributor)
Created2025-05
Description
We measure non-parametric morphological statistics, including Concentration (C), Asymmetry (A), and Smoothness (S), as well as the Gini coefficient and M20 moment of light, using Statmorph. Additionally, we analyze the variation of these parameters as a function of wavelength and radius to assess structural differences across different stellar populations. To place PEARLSDG in the broader context of dwarf galaxy morphology, we compare its structural parameters to a sample of 30 dwarf irregular galaxies (van Zee 2000, Conselice 2003). Furthermore, we investigate the presence of isophotal twisting, measuring orientation angle as a function of radius, to explore potential evidence of past dynamical interactions.
Our results aim to clarify whether PEARLSDG’s quiescence and structure arise from internal evolutionary processes or past interactions with neighboring galaxies. These findings contribute to our
understanding of dwarf galaxy morphology and the environmental factors that shape their evolution.
ContributorsMcLeod, Noah (Author) / Jansen, Rolf (Thesis director) / Carleton, Timothy (Committee member) / Windhorst, Rogier (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor) / School of Mathematical and Statistical Sciences (Contributor) / School of Earth and Space Exploration (Contributor)
Created2025-05
DescriptionThis project focuses on creating a detour recommendation system as a component of a larger product. The criteria for detours in this implementation are interesting locations that can be enjoyed for free. The implementation utilizes API calls, web scraping, and machine learning to determine if a location qualifies a detour.
ContributorsYang, Nathan (Author) / Chavez-Echeagaray, Maria Elena (Thesis director) / Mclaughlin , Ranique (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2025-05