Computer Engineering Commons^™

Regional Sea Level Rise Prediction In Monterey Bay With Lstms And Vertical Land Motion, Branden Lopez

Master's Projects

Earth system data is vast in volume and variety, and is used to forecast weather,

hurricanes, floods, and sea level. Sea Level Rise (SLR) impacts various sectors, espe- cially ecosystems, food production, industry, population, health, and the availability of

clean water. Because of its broad impact, describing the behavior and forecasting SLR is an important topic. Traditional Machine Learning (ML) models vary in use, but many are not capable of capturing all the non-linear spatial and temporal properties of SLR factors. Deep learning models efficaciously handle complex time series data, noise, and high dimensional spaces, making them a focus of …

Distinguishing Chatbot From Human, Gauri Anil Godghase

Master's Projects

There have been many recent advances in the field of Generative Artificial Intelligence and Large Language Models, with GPT 3 or ChatGPT model being one of the frontrunners in this field. These large language models have become so powerful that it has become difficult to differentiate between text written by humans and machine-generated text. This paper proposes a solution to the problem of classification of the origin of data (human or chatbot) by using Machine Learning. In addition, the proposed solution also helps us analyze the text generated by such Language Models and understand the underlying patterns present in the …

Unveiling Gender Bias: An Eye-Tracking Analysis Of Scene Perception, Naga Srija Gopisetty

Master's Projects

Gender bias deeply affects how we perceive and interact with the world around us. This study examined how gender bias affects scene perception by using eye-tracking technology. The relation between gender bias and pupil dilation is examined utilizing art as a stimuli and also studied the underlying cognitive processes. This experiment combines image presentations along with input from participants, drawing on previous research in gender bias, cognitive psychology, and eye-tracking methodologies. The experiment design showcases a series of images to participants, subtly replacing one image during the trial, and then asks participants whether the

replacement took place. This study uses …

Exploring Gender Bias In Large Language Models: Cross-Linguistic Comparisons And Evaluation Letters Analysis, Athira Kumar

Master's Projects

Large language models (LLMs) play a significant role in modern human-computer interaction. They have exploded in popularity recently, becoming widely used for various tasks. However, concerns persist regarding potential biases within these models. This project investigates gender bias in the popular LLMs - GPT-3.5, GPT-4, Gemini, and LLAMA. The first part of our study focuses on analyzing biases using ambiguous sentences across three languages - English, Malayalam, and Tamil. We evaluate the LLMs to see if they associate occupations with commonly held gender stereotypes, by using specific professions within our test sentences. Through the use of two low-resource languages, this …

Deception Detection Models From Speech, Tien Nguyen

Master's Projects

Recently, researchers have shown an increased interest in automatically detecting deceptive actions. The attention given to this area can be attributed to the many potential applications of deception detection, especially in the field of criminology. To contribute to the deception detection research, this project investigates textual and audio data extracted from spoken and written words. We evaluated and compared the traditional linguistic models with advanced Large Language Models (LLMs) while using Natural Language Processing (NLP) techniques. Additionally, various feature selection techniques were applied to assess the importance of linguistic features. We conducted extensive experiments to evaluate the effectiveness of both …

Hindi Image Captioning Using Indictrans2 And Encoder-Decoder Architecture, Anahita Vayalombrone Dinesh

Master's Projects

One of the most prominent tasks that lie on the conjunction of Natural Language Processing (NLP) and computer vision, is image captioning. Image captioning is the generative task of achieving textual descriptions from images. Its application finds use in many real-world scenarios like aiding the visually impaired, editing applications, recommendation systems, and medical imaging. This research focus lies in Hindi image captioning, the official language of India, as it has not been explored as far as its need. Several challenges such as the lack of substantial Hindi text data for training models, the need for human annotators to verify the …

Lexigen: Lexical-Driven Image Generation, Sangram Prashant Chincholkar

Master's Projects

This research project proposes a novel approach to user-driven image editing via natural language descriptions. The aim is an accurate change of certain features of an image with respect to the descriptive text while maintaining, with equal concern, the integrity of the remaining parts of the image not affected by the description. The task is particularly relevant for fields like content creation, personalized design, and automated image editing that require both coherence of a visual scene and textual description. We propose a generative model, LexiGen, which perfectly integrates natural language descriptions with their corresponding visual changes within an image. The …

Comparison Of Protein Structures Predicted By Genai Tools In A Zero-Shot Manner, Kruthi Shankar Rao

Master's Projects

Generative AI models have vast applications and one such critical application explored in this study is protein structure prediction. The 3D structures of proteins determine their function. Our study mainly focuses on using generative AI models such as ESMFold and ColabFold to predict and examine naturally occurring and mutated sequences. The workflow begins with collecting antimicrobial resistance (AMR) and toxin-antitoxin (TA) protein data. The sequences are applied over pretrained AI models to predict protein structures. Following this, models are fine-tuned with original and mutated target datasets. A comparison of models’ performances is done using metrics such as root mean square …

Exploring The Use And Misuse Of Large Language Models (Llms), Hezekiah Paul D. Valdez

Master's Projects

Large Language Models (LLMs) have quickly gone from simple rule-based systems to complex knowledge bases capable of tackling many different tasks across a variety of fields. What began as an exercise in human-computer interaction has become the basis for artificial intelligence in a variety of mediums. When attached to larger systems, LLMs become generative assistants that can perform highly on human proficiency assessments and other benchmark skill assessments. This increase in proficiency has led these systems to be deployed in fields such as cybersecurity, business, and programming to help improve productivity and efficiency. However, such a wide availability has allowed …

Enhancing Qwen2.5-Coder: A Deep Dive Into Fine-Tuning Using Peft For Superior Code Outputs, Lohith Nagaraja

Master's Projects

The main objective of this research is to improve the quality of software code that is produced by the Qwen2.5-Coder model specifically in terms of maintainability, complexity, and reliability. Our approach is going to be a more specific one that will involve the Parameter-Efficient Fine Tuning (PEFT) framework combined with quantization through Low-Rank Adaption (LoRA). This approach involves fine-tuning only some of the parameters of a model to make it suitable for software programming with the general structure of the model largely intact. In this paper, SonarQube is used as a tool to help quantify the improvements made to the …

Llamatalk: Empowering Conversations With Retrieval-Augmented Generation, Aravind Rokkam

Master's Projects

This research report talks about the implementation and a comparative study of Llama 7B model’s fine-tuning technique and Retrieval Augmented Generation (RAG) capabilities in the context of creating a reliable AI therapist. This study focuses on training these models using diverse datasets consisting of doctor-patient conversations predominantly addressing general health issues. Using a technique like fine-tuning within the Llama 7B model, the project focuses on training the model with a diverse dataset comprising doctor-patient interactions primarily addressing general health concerns. Additionally, carefully organized mental health dataset from HOPE dataset, ensuring the bot's responsiveness to mental health inquiries. Through integration with …

Wall-E: An Autonomous Ai Rover For Precision Agriculture, Simar Ghumman

Master's Projects

Unmanned Ground Vehicles (UGVs) are emerging as a crucial tool in the world of precision agriculture. By working with UGVs equipped with machine learning, we can find solutions to a range of complex agricultural problems. My project, titled “Wall-E: Artificial Intelligence Robot for Precision Agriculture,” focuses on developing a UGV capable of navigating through agriculture fields autonomously while capturing data. Using machine learning, computer vision, and other sensor technologies, Wall-E is capable of estimating the total yield of crops, self-localization, mapping its environment in real time, and avoiding obstacles along its route. The purpose of this project is to automate …

Adaptive Metric-Driven Load Balancing For Specialized Clusters Using Nginx, Juhi Raju Malkani

Master's Projects

Adaptive Metric-Driven Load Balancer is an innovative two-tier load-balancing system that uses NGINX and Prometheus to optimize resource allocation in specialized cloud clusters. This framework is built to give great performance and flexibility and runs on Google Kubernetes Engine (GKE), but it may also be deployed on local cloud environments for added security. The first tier of our system uses an NGINX-based load balancer to route incoming requests based on content type, sending traffic to hardware-optimized clusters for processing requests through specialized hardware. In our algorithm, the second tier dynamically modifies load distribution throughout each cluster by calculating pod weights …

Gradual Typing For Information Flow Control In Typescript Using Es Lint, Ashish Agarwal

Master's Projects

Current state-of-the-art systems tackle data security threats by incorporating information flow control (IFC) to ensure that a piece of information reaches only its intended recipient. However, most IFC implementations introduce a custom language built on top of a well-known language. Adaptations of such languages are limited due to limited support and updates, along with difficulty in learning new syntaxes. Implementations without a custom language offer incomplete IFC support. We present a comprehensive framework by leveraging Typescript, in conjunction with ESLint and NodeJS, aiming to resolve some of the limitations of IFC and intending to facilitate acceptance by a wide range …

Machine Learning Applications And Sustainable Development, Vishnu Pendyala

Open Educational Resources

This presentation, "Machine Learning Applications and Sustainable Development," explores the intersection of machine learning and its impact on privacy, equity, and societal well-being. It delves into the potential for re-identification of "anonymized" data through various techniques like k-anonymity, L-diversity, and the vulnerabilities of large language models, illustrating these concepts with real-world examples such as the AOL search data, Netflix Prize dataset, and Strava's fitness tracking. The presentation also discusses solutions to enhance data privacy, including differential privacy and the emerging field of machine unlearning, highlighting their applications and limitations. Finally, it addresses the broader implications for civil rights and ethical …

Examining The Externalities Of Highway Capacity Expansions In California: An Analysis Of Land Use And Land Cover (Lulc) Using Remote Sensing Technology, Serena E. Alexander, Bo Yang, Owen Hussey, Derek Hicks

Mineta Transportation Institute

There are over 590,000 bridges dispersed across the roadway network that stretches across the United States alone. Each bridge with a length of 20 feet or greater must be inspected at least once every 24 months, according to the Federal Highway Act (FHWA) of 1968. This research developed an artificial intelligence (AI)-based framework for bridge and road inspection using drones with multiple sensors collecting capabilities. It is not sufficient to conduct inspections of bridges and roads using cameras alone, so the research team utilized an infrared (IR) camera along with a high-resolution optical camera. In many instances, the IR camera …

Electrical Vehicle Charging Infrastructure Design And Operations, Yu Yang, Hen-Geul Yeh

Mineta Transportation Institute

California aims to achieve five million zero-emission vehicles (ZEVs) on the road by 2030 and 250,000 electrical vehicle (EV) charging stations by 2025. To reduce barriers in this process, the research team developed a simulation-based system for EV charging infrastructure design and operations. The increasing power demand due to the growing EV market requires advanced charging infrastructures and operating strategies. This study will deliver two modules in charging station design and operations, including a vehicle charging schedule and an infrastructure planning module for the solar-powered charging station. The objectives are to increase customers’ satisfaction, reduce the power grid burden, and …

Slowly Changing Dimensions And Fast Changing Facts - The Story Of The Traditional Data Warehouse, Vishnu Pendyala

Open Educational Resources

This presentation, "Slowly Changing Dimensions and Fast Changing Facts - The Story of the Traditional Data Warehouse," explores how to generate actionable insights from organizational performance data. It begins by discussing the challenges of generating insights directly from existing databases and introduces the concept of data cubes for efficient aggregate storage. The core of the presentation focuses on dimensional modeling, particularly the Star Schema, as a solution for analytical processing needs that cannot be met by transactional systems. It details the characteristics of fast-changing fact tables (long, lean, numeric) and slowly changing dimension tables (descriptive, wide, containing strings and numbers). …

The Fashion Visual Search Using Deep Learning Approach, Smita V. Bhoir, Sunita R. Patil

Library Philosophy and Practice (e-journal)

In recent years, the World Wide Web (WWW) has established itself as a popular source of information. Using an effective approach to investigate the vast amount of information available on the internet is essential if we are to make the most of the resources available. Visual data cannot be indexed using text-based indexing algorithms because it is significantly larger and more complex than text. Content-Based Image Retrieval, as a result, has gained widespread attention among the scientific community (CBIR). Input into a CBIR system that is dependent on visible features of the user's input image at a low level is …

Chatgpt As Metamorphosis Designer For The Future Of Artificial Intelligence (Ai): A Conceptual Investigation, Amarjit Kumar Singh (Library Assistant), Dr. Pankaj Mathur (Deputy Librarian)

Library Philosophy and Practice (e-journal)

Abstract

Purpose: The purpose of this research paper is to explore ChatGPT’s potential as an innovative designer tool for the future development of artificial intelligence. Specifically, this conceptual investigation aims to analyze ChatGPT’s capabilities as a tool for designing and developing near about human intelligent systems for futuristic used and developed in the field of Artificial Intelligence (AI). Also with the helps of this paper, researchers are analyzed the strengths and weaknesses of ChatGPT as a tool, and identify possible areas for improvement in its development and implementation. This investigation focused on the various features and functions of ChatGPT that …

A Natural Language Processing Approach To Malware Classification, Ritik Mehta

Master's Projects

Many different machine learning and deep learning techniques have been successfully employed for malware detection and classification. Examples of popular learning techniques in the malware domain include Hidden Markov Models (HMM), Random Forests (RF), Convolutional Neural Networks (CNN), Support Vector Machines (SVM), and Recurrent Neural Networks (RNN) such as Long Short-Term Memory (LSTM) networks. In this research, we consider a hybrid architecture, where HMMs are trained on opcode sequences, and the resulting hidden states of these trained HMMs are used as feature vectors in various classifiers. In this context, extracting the HMM hidden state sequences can be viewed as a …

Deep Learning In Ai Medical Imaging For Stroke Diagnosis, James Mario Guzman

Master's Theses

Enhancing medical imaging stroke diagnosis applications with artificial intelligence (AI) tools to determine lesion volume, location and clinical metadata is vital toward guiding patient treatment and procedure. A major hardship in developing stroke diagnosis AI tools is the scarcity of publicly available clinical 3D stroke datasets. Through working with Johns Hopkins University, University of Michigan’s ICPSR data repository and SJSU research, we gained access to potentially the largest 3D MRI stroke dataset with clinical metadata annotated by neuroradiologists known as ICPSR 38464. With the ICPSR 38464 dataset recently being available through institutional review board (IRB) approval or exemption, we were …

Image-Based Classification Of Malware Using T-Sne Images, Vincent Stowbunenko

Master's Projects

This Master’s project proposes a novel technique for classifying malware using image-based methods. The approach involves generating t-SNE images from the EMBER dataset, which contains one million samples of both malware and benign files, each represented by over 2,000 features. The t-SNE technique is well-suited for capturing intricate patterns in complex datasets because it effectively maintains the local structure. These t-SNE images are then used as inputs to train two lightweight image classification models, SqueezeNet and MobileNet. Additionally, to provide a benchmark for comparison, a non-image classification model using LightGBM is also explored.

As part of the investigation, the project …

Malware Classification Using Opcode N-Grams And Word Embeddings, Siddhita Joshi

Master's Projects

Malware is a serious risk to any software application whether it is standalone or over the network. In order to protect computer systems, it is essential to detect and classify malware effectively. Modern malware classification research focuses on Machine Learning and Deep Learning techniques to identify advanced malicious software. This project explores malware classification by combining two robust methods: n-grams and word embedding. By extracting opcode n-grams, we make use of sequential nature of malware execution to identify any local patterns within the malware executable.

We use word embedding methods such as Word2Vec, Doc2Vec, and FastText to produce dense vector …

Evaluation Of The Effect Of Walnut Extract On Sp1-Related Pathways, Jihan Yehia

Master's Projects

Walnut extract (WE) has shown promising anti-cancer effects, such as inducing apoptosis and moderating cell cycle progression. A previous study by Dr. Brandon White’s Lab at San Jose State University hypothesizes that WE can downregulate the expression of the pro-tumoral specificity protein 1 (Sp1) in triple negative breast cancer (TNBC). This project builds an RNA-seq pipeline that runs differential gene expression (DGE) analysis to study the effect of WE on TNBC, thereby offering a wider perspective on genes that may be affected by this treatment. The data used in this project originated from Illumina and Nanopore sequencing methods, and DGE …

Efficient Video Qoe Prediction In Intelligent O-Rans, Aditya Kulkarni

Master's Projects

Open Radio Access Network (O-RAN) is a platform developed by a collaboration between wireless operators, infrastructure vendors, and service providers for deploying mobile fronthaul and midhaul networks, built entirely on cloud-native principles. The vision of O-RAN lies in the virtualization of traditional wireless infrastructure components, like Central Units (CU), Radio Units (RU), and Distributed Units (DU). O-RAN decouples the above-mentioned wireless infrastructure components into opensource elements, operating consistently with other elements of different vendors in the network. Quality of Experience (QoE) deals with a user’s subjective measure of satisfaction. RAN Intelligent Controller (RIC) in O-RAN provides flexibility to intelligently program …

Implicit Racial Bias: A Human Computer Interaction Study Using Eye Tracker, Wenfan Zhang

Master's Projects

Contemporary Human-Computer Interaction (HCI) research has an increasing emphasis on reducing ethnicity bias. The study presents a new method to explore and reduce biases using detailed experiments. The experimental procedure involves presenting participants with images of ethnically diverse characters across three conditions. The study's results significantly illuminate ethnicity bias in character selection dynamics. Participants exposed to targeted training interventions displayed a significant shift in preferences for characters engaged in intellectual activities. Notably, this shift was influenced by the ethnicity of the characters involved. Interestingly, the eye-tracking data unveiled distinct patterns of cognitive load, characterized by slower response times and greater …

Enhancing Driver Distraction Detection Through The Synergy Of Deep And Traditional Machine Learning, Gowtham Chandrasekaran

Master's Projects

Distracted driving is a major contributor to motor vehicle accidents, causing injury and loss of life. It is one of the major factors that affect the overall driving behavior of a person. Insurance companies take into consideration factors like gender, age, etc. to set insurance premiums for their customers. Today, machine learning and artificial intelligence can eradicate this bias. A machine learning model can analyze driving behavior, such as the frequency and severity of accidents, the speed at which they drive, and their habits such as distracted driving. Based on this information, the model can then determine the risk of …

Unlearning Hidden Bias Between Refugees : An Initial Empirical Investigation, Akshay Sunil Gurnaney

Master's Projects

The challenges that refugees in various regions encounter are common knowledge. One such challenge is a bias among refugees on ethnocentric grounds. In particular, there are various articles that have pointed out the struggles faced by Syrian refugees in countries like Europe as a result of implicit bias. In fact, the media coverage of Syrian crises and the government responses to the same shed negative light on the refugees themselves. On the contrary, the media coverage of Ukrainian crises is very different with lesser restrictions from the governments.

This research attempts to identify the extent of implicit bias between Ukrainian …

Poriferal Vision: Using Mobilenet To Classify Sponge Spicules Through Transfer Learning, Brian Tran

Master's Projects

Global warming is an ongoing issue where the Earth is rapidly warming up. It negatively affects the growth of coral through ocean warming and ocean acidification. Many coral communities, home to a large variety of marine life, are expected to be severely impacted by these effects. Past evidence suggests that sponges will take over as the primary reef builders since many species of sponges have skeletons made of silica or glass which is not affected by ocean acidification. More research is needed to determine which kinds of sponge will most likely be able to thrive in today’s climate.

This can …