Digital Commons Network^™

Steganographic Capacity Of Selected Machine Learning And Deep Learning Models, Lei Zhang

Master's Projects

As machine learning and deep learning models become ubiquitous, it is inevitable that there will be attempts to exploit such models in various attack scenarios. For example, in a steganographic based attack, information would be hidden in a learning model, which might then be used to gain unauthorized access to a computer, or for other malicious purposes. In this research, we determine the steganographic capacity of various classic machine learning and deep learning models. Specifically, we determine the number of low-order bits of the trained parameters of a given model that can be altered without significantly affecting the performance of …

Leveraging Tweets For Rapid Disaster Response Using Bert-Bilstm-Cnn Model, Satya Pranavi Manthena

Master's Projects

Digital networking sites such as Twitter give a global platform for users to discuss and express their own experiences with others. People frequently use social media to share their daily experiences, local news, and activities with others. Many rescue services and agencies frequently monitor this sort of data to identify crises and limit the danger of loss of life. During a natural catastrophe, many tweets are made in reference to the tragedy, making it a hot topic on Twitter. Tweets containing natural disaster phrases but do not discuss the event itself are not informational and should be labeled as non-disaster …

Federated Learning For Protecting Medical Data Privacy, Abhishek Reddy Punreddy

Master's Projects

Deep learning is one of the most advanced machine learning techniques, and its prominence has increased in recent years. Language processing, predictions in medical research and pattern recognition are few of the numerous fields in which it is widely utilized. Numerous modern medical applications benefit greatly from the implementation of machine learning (ML) models and the disruptive innovations in the entire modern health care system. It is extensively used for constructing accurate and robust statistical models from large volumes of medical data collected from a variety of sources in contemporary healthcare systems [1]. Due to privacy concerns that restrict access …

Keystroke Dynamics And User Identification, Atharva Sharma

Master's Projects

We consider the potential of keystroke dynamics for user identification and authentication. We work with a fixed-text dataset, and focus on clustering users based on the difficulty of distinguishing their typing characteristics. After obtaining a confusion matrix, we cluster users into different levels of classification difficulty based on their typing patterns. Our goal is to create meaningful clusters that enable us to apply appropriate authentication methods to specific user clusters, resulting in an optimized balance between security and efficiency. We use a novel feature engineering method that generates image-like features from keystrokes and employ multiclass Convolutional Neural Networks (CNNs) to …

Modeling Sequencing Artifacts In Artificial Low Frequency Cancer Data, Hannele Padre

Master's Projects

The rapid advancement in technology for next-generation sequencing (NGS) continues to make NGS more affordable, and in turn, there is a high influx of sequencing data. While NGS is relatively fast and efficient compared to previous sequencing technologies, there are a multitude of steps in the NGS workflow in which sequencing errors can be introduced. Such sequencing errors are known as artifacts, and if not careful, they can be mistaken for true variants. It's especially important to distinguish artifacts in cancer biopsies, specifically liquid biopsies, a noninvasive method for sample collection. Somatic mutations occur at low frequencies, and a liquid …

The Search For Metabolic Variants In Response To Climate Change In The American Pika, Tyler Stewart Trader

Master's Projects

Climate change and rising temperatures pose a serious threat to the long-term survival of American pika (Ochotana princeps), emphasizing the interest in the adaptive capability of the pika. This project queried single nucleotide polymorphisms in a population of American pika in Yosemite National Park using Whole Genome Sequencing data, with a specific interest in metabolic variants. The sample data included temporally separated cohorts, comparing modern population data to historical data taken before rapid anthropogenic climate change. Statistically significant variants were identified under Approximate Bayesian Computation using a population decline model. Although population statistics indicated little change between the …

Identification Of Oral Microbiome With Full-Length 16s Sequencing Using Nanopore Technology, Uyen Hoang

Master's Projects

Human microbiota has a key role in human development and receives considerable attention from research communities. Advances in metagenomics sequencing technologies have significantly improved the capabilities to characterize microbial species associated with human health and disease. Nanopore metagenomics sequencing makes a revolution by detailed characterization of the complex human microbiome. The Nanopore approach resembles traditional genome sequencing but also presents unprecedented challenges. This project will examine the Nanopore approach for oral human microbiome analysis focusing on the 16S rRNA gene collected from the oral cavity of healthy and Mild PKU patients. The findings of this project found that Nanopore sequencing …

Machine Learning-Based Anomaly Detection In Cloud Virtual Machine Resource Usage, Tarun Mourya Satveli

Master's Projects

Anomaly detection is an important activity in cloud computing systems because it aids in the identification of odd behaviours or actions that may result in software glitch, security breaches, and performance difficulties. Detecting aberrant resource utilization trends in virtual machines is a typical application of anomaly detection in cloud computing (VMs). Currently, the most serious cyber threat is distributed denial-of-service attacks. The afflicted server's resources and internet traffic resources, such as bandwidth and buffer size, are slowed down by restricting the server's capacity to give resources to legitimate customers.

To recognize attacks and common occurrences, machine learning techniques such as …

Cultural Perspectives In Pain And Palliative Care, Loan Lagué

Master's Projects

Culturally and linguistically diverse (CALD) patients are constantly challenged by the different levels on which lack of cultural competency or bias occurs in healthcare. This literature review focuses on the healthcare providers’ experiences when providing pain and palliative care to CALD populations and explores the challenges that may result in health inequalities. A review of current literature was conducted using PubMed, CINHAL, Sage Journals, Ovid, and Science Direct electronic databases. Sixteen published articles between 2015-2023 were reviewed. Findings from this review identified factors that induce poor quality palliative, pain, and end-of-life (EOL) care among culturally diverse groups. Improving quality of …

Standardizing Patient Transfer Process Among Nurses From Labor & Delivery To Mother-Baby Unit: A Quality Improvement Project, Kiratraj Kaur Grewal

Master's Projects

Purpose: To implement and evaluate the efficacy of a standardized patient transfer process among registered nurses (RN) from Labor & Delivery (L&D) to Mother-Baby Unit (MBU) using PDSA cycle to improve RN satisfaction and reduce patient transfer time
Conceptual Framework: PDSA Cycle
Setting: Labor & Delivery and Mother-Baby Unit of a tertiary care hospital
Methods: The project consisted of collecting both objective (estimated patient transfer time from L&D to MBU) and subjective data (anonymous nursing pre- and post-intervention surveys). Standardization of transfer process included implementing a transfer checklist, patient transfer tool, updating the workflow, and a sample patient transfer video. …

A Roadmap To Safer Hyaluronic Acid Injections, Mandi Lonergan

Master's Projects

Non surgical aesthetic treatments have become increasingly more popular in the United States (2020 Plastic Surgery Statistics Report 2021). Hyaluronic acid (HA) fillers are the second most popular non-surgical aesthetic treatment behind Botulinum toxin injections (The Aesthetic Society, 2021). The increased use of these fillers has led to increased rates of complications With the release of new long lasting fillers in 2014, providers began to see an increase in delayed onset nodules (DON). A DON is considered in the aesthetic field to be a nodule that occurs 30 days to years after filler injection (Convery et al., 2021). A …

Real Time Panoramic Image Processing, Matthew Gerlits

Master's Projects

Image stitching algorithms are able to join sets of images together and provide a wider field of a vision when compared with an image from a single standard camera. Traditional techniques for accomplishing this are able to adequately produce a stitch for a static set of images, but suffer when differing lighting conditions exist between the two images. Additionally, traditional techniques suffer from processing times that are too slow for real time use cases. We propose a solution which resolves the issues encountered by traditional image stitching techniques. To resolve the issues with lighting difference, two blending schemes have been …

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 …

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 …

On The Memory Scalability Of Spectral Clustering Algorithms, Ran Li

Master's Projects

Spectral clustering has lots of advantages compared to previous more traditional clustering methods, such as k-means and Gaussian Mixture Models (GMM), and is popular since it was introduced. However, there are two major challenges, speed scalability and memory scalability, that impede the wide applications of spectral clustering. The first challenge has been addressed recently by Chen [1] [2] in the special setting of sparse or low dimensional data sets. In this work, we will first review the recent study by Chen that speeds up spectral clustering. Then we will propose three new computational methods for the same special setting of …

Metagenomic Survey Of Marine 16s Bacterial Communities Off Palmer Station In Antarctica, Daniel Salter

Master's Projects

This project surveys the metagenomic bacterial community composition in marine surface waters off Palmer Station, Western Antarctic Peninsula and correlates findings with temperature and salinity data. Marine bacterial communities play a vital role in nutrient cycling, but data on surface waters in this region are limited. Analyzing fifteen samples of 16S sequencing data from three austral summers, consistent dominance was observed by the classes Alphaproteobacteria, Gammaproteobacteria, and Flavobacteria. Correlation analysis confirmed significant relationships between taxa and environmental conditions. The observed trends suggest varying abilities of phyla to resist and adapt to changing environmental conditions. Notably, Alphaproteobacteria demonstrated adaptability to favorable …

Gender Classification Via Human Joints Using Convolutional Neural Network, Cheng-En Sung

Master's Projects

With the growing demand for gender-related data on diverse applications, including security systems for ascertaining an individual’s identity for border crossing, as well as marketing purposes of digging the potential customer and tailoring special discounts for them, gender classification has become an essential task within the field of computer vision and deep learning. There has been extensive research conducted on classifying human gender using facial expression, exterior appearance (e.g., hair, clothes), or gait movement. However, within the scope of our research, none have specifically focused gender classification on two-dimensional body joints. Knowing this, we believe that a new prediction pipeline …

Dynamic Predictions Of Thermal Heating And Cooling Of Silicon Wafer, Hitesh Kumar

Master's Projects

Neural Networks are now emerging in every industry. All the industries are trying their best to exploit the benefits of neural networks and deep learning to make predictions or simulate their ongoing process with the use of their generated data. The purpose of this report is to study the heating pattern of a silicon wafer and make predictions using various machine learning techniques. The heating of the silicon wafer involves various factors ranging from number of lamps, wafer properties and points taken in consideration to capture the heating temperature. This process involves dynamic inputs which facilitates the heating of the …

Sending And Receiving Internet Messages From Disconnected Areas, Abhishek Prakash Gaikwad

Master's Projects

Over 62% of the world is connected to the internet with more than 6.9 billion smartphone users. The omnipresence of technology in the form of the internet and smartphones have led to constant research in improving communication throughout the world. But even today, 37% (2.9 billion people) are not connected to the internet even though most of the people in such areas have smartphones. To solve this problem of access to internet services in disconnected areas, a software-only mobile-first approach has been proposed for disconnected data distribution infrastructure which can support different internet applications in limited connectivity. A prototype application …

Deep Learning Neural Machine Translation Conversational Agent, Abhishek Vaid

Master's Projects

Neural Machine Translation (NMT) is a prominent natural language processing technique that is being used to develop conversational AI technology. However, most chatbots do not provide live API features and have list-based scripted responses. Most chatbots are majorly restricted by the training data on which they were trained on and have no knowledge of current events. This research project intends to research and develop an approach to providing live information. We experiment with various techniques in terms of the type of data being used to harness live capabilities. We optimize the hyperparameters that are needed for a Conversational AI agent …

Job Tailored Resume Content Generation, Sumedh Kale

Master's Projects

Generally candidates apply to multiple jobs with a single resume and do not tend to customize their resume to match the job description. This hampers their chances of getting a resume shortlisted for the job. The project aims to help such candidates build job tailored resumes that help them create a customized and targeted resume for a specific job or industry. The tool specifically targets candidates’ employment history, for resume content generation. We then use natural language processing

(NLP) techniques to extract and organize this data into a structured format for the dataset. We experiment with multiple variations of the …

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 …

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 …

Image-Based Malware Detection Using Convolutional Neural Network Techniques, Brandon Palomino

Master's Projects

In this study, we delve into the realm of malware detection and classification, leveraging the capabilities of different Convolutional Neural Networks (CNNs). Our approach involves transforming executable files into image formats and subsequently applying advanced CNN techniques for image recognition. The study emphasizes the use of two distinct CNN architectures: a traditional CNN model and a modified CNN variant known as a convolutional recurrent neural network (CRNN), each with unique structural and functional attributes. To effectively train these models, we adopt a transfer learning strategy, utilizing pre-existing CNN models that have been extensively trained on large-scale image datasets. This methodology …

Advances In Robustness Of Image-Based Malware Detection, Rishika Pamanji

Master's Projects

In recent years, deep learning has emerged as a powerful tool for image classification tasks. However, the performance of individual deep learning models can be limited by their architecture and training data. In this project, various Convolutional Neural Network (CNN) architectures are proposed to train the malware data for feature extraction for various color coordinates such as L, CMYK, RGB, RGBA, and YCbCr. Different optimization techniques like Stochastic Gradient Descent, Root Mean Square Propagation, Ada Delta, Adam, and Adaptive Gradient are used to minimize errors in the trained data, leading to enhanced accuracy. The proposed ensemble deep learning model for …

Spartandark: Anonymity Model Integration With A Blockchain Network Using Spartangold, Nishanth Uchil

Master's Projects

Demand for blockchain ecosystems has seen exponential growth in recent times due to its decentralized nature and trustless verification process for the transactions involved. However, transaction data needs to be leveraged for verification, which coupled with the transparent nature of the blockchain ledger, provides sufficient data for malicious entities to reveal identities and even financial history of users. Data masking techniques have been employed over the years to make blockchain transactions anonymous, making them resistant to identity analysis, a key set of methods being zero-knowledge proof (zk-proof) protocols that guarantee zero data leak. In this research, we develop SpartanDark, a …

Enhancing The Queueing Process For Yioop's Scheduler, Gargi Sheguri

Master's Projects

Indexing in search engines is the process of storing information related to crawled pages to facilitate searches. A crucial determinant of the success of a search engine is the efficiency of the indexing process utilized, which greatly affects both the speed and relevancy of search results. Yioop is an open-source web search engine that employs an inverted index strategy, wherein each term is mapped to a list of the documents it appeared in while crawling.

The primary aim of this project is to better the indexing system used by Yioop, and thus improve the quality of the Search Engine Results …

Visual Scene Classification Using Ensemble Of Machine Learning Classifiers, Rahul Ranganath

Master's Projects

Visual scenes represent the comprehensive visual information observed in a particular environment. Whether natural landscapes, urban settings, or designed interiors, visual scenes encompass the arrangement of elements that individuals perceive through their visual senses. Visual search is perhaps one of the most typical jobs that we carry out several times a day. This is one of the main paradigms for researching visual attention. Many visual task models have been put forward in an effort to better understand visual attention. Fixations and the rapid movement of the eye - saccades, define visual exploration and visual search. When we subject viewers to …

Interest-Based Recommendation System Using Gmail Topic Modelling, Pranav Ghaskadbi

Master's Projects

Emails are a fundamental part of modern communication. Much of communicative discourse in modern society occurs over email, resulting in personal collections for each mail user which are rich in latent user’s interests. Conventional recommendation systems require historical data of user activity and interactions to derive user interests. The absence of activity and interaction data poses an interesting challenge for generating relevant recommendations for users. We were motivated to investigate approaches to identify user interests in the absence of historical data to generate personalized content recommendations. There is opportunity to derive user interests from email data, which can be used …