Digital Commons @ Harrisburg University of Science and Technology
Not a member yet
435 research outputs found
Sort by
Vermifilter Treatment of Aquaculture Effluent
No full textThis research characterizes the degree of solids reduction, nutrient mineralization, and organic removal in aquaculture effluent using vermifilter reactors for reuse as a hydroponic nutrient solution. Success would allow the development of a circular nutrient economy and hydroponic industries to support a closed-loop, integrated farming system for increased food security. (Independent Research
Decorrelated Deep Neural Networks: Learning Bias Invariant & Scanner Independent Features, and Causal relationships Using a novel deep learning methods based on Distance Correlation
Get PDFAdvancements in deep learning or deep neural networks have made it possible to reach expert-level performance in a variety of applications, even in challenging situations. However, a central challenge in all deep learning, as well as machine learning applications, is dealing with its dependency on the quality of data which can be significantly impacted by biases, confounders, and irrelevant variations in data which leads to spurious relationships and erroneous decisions. The main purpose of this dissertation is to build a robust deep learning model which considers and mitigates these biases. Another challenge with the deep learning model is learning associations present in the data rather than causations. This also leads to bias problems and non-interpretable systems. So, the purpose of this dissertation also includes introducing causality in the deep learning models. Thus, developing a novel deep learning model to learn bias invariant features and learn causal discoveries are promising areas of research with high potential impact.
The overarching goal of this dissertation is to improve the performance, reliability, and generalization ability of deep learning even in the presence of biases and spurious associations in the data. This entails several research directions. First, we introduce a decorrelated framework that addresses the imbalanced and scanner dependencies issues present in the Parkinson’s Disease (PD) dataset. Second, we further define the general formulation of decorrelated deep learning models. This provides the foundation for generic bias mitigation analysis and the design of robust decorrelated deep learning models. This dissertation also focuses on the topic of Granger Causality (GC) introduction in the deep learning model. Thus, the third research direction includes extending the LSTM-based Granger Causality framework to incorporate Graph Neural Network (GNN) and distance correlation which enables improvement in the performance of the deep learning model and provides interpretable GC interactions.
We propose a novel bias mitigation method for deep learning models by leveraging the distance correlation function to decorrelate the features and biases to provide a robust solution. We explore the use of this method in neuroimaging study settings for disease classification. We also derive the generic decorrelation-based bias mitigation framework for different data scenarios and different deep learning architectures. These results show how our approach provides a robust, flexible, scalable, and generic framework that improves the performance of deep learning models while reducing bias effects on model predictions. In addition to this, we define a mathematical framework to introduce the fusion of GC with GNN and distance correlation and showcase their success in learning complex non-linear Granger causal connections. We study the implications of our work in the deep learning field and discuss future work to further leverage this robust decorrelated framework and improve the performance irrespective of the quality of data
Blockchain Storage – Drive Configurations and Performance Analysis
Get PDFThis project will analyze the results of trials implementing various storage methods on Geth nodes to synchronize and maintain a full-archive state of the Ethereum blockchain. The purpose of these trials is to gain deeper insight to the process of lowering cost and increasing efficiency of blockchain storage using available technologies, analyzing results of various storage drives under similar conditions. It provides performance analysis and describes performance of each trial in relation to the others
Abscond Runner
Get PDFAbscond Runner is a game we\u27ve made based on the movie and book series The Maze Runner . This is a first-person novel experience where the player must go through the maze by maneuvering through obstacles and destroying the enemies to reach the end. (Class Project
Favorite Programming Language among Students
Get PDFThis project involves understanding the favorite programming language among students. I hypothesize that the favorite programming language will be Python. (Class Project
Cloud Container Security’ Next Move
Get PDFIn the last few years, it is apparent to cybersecurity experts everywhere that the proverbial container tech genie is out of the bottle, and has been widely embraced across multiple organizations. To achieve the flexibility of building and deploying applications anywhere and everywhere, cloud native environments have gained great momentum and made the development lifecycle simpler than ever. However, container environments brings with them a range of cybersecurity issues that includes images, containers, hosts, runtimes, registries, and orchestration platforms, which needs the necessity to focus on investing in securing your container stack.
According to this report[1], released by cloud-native security provider Aqua Security on June 21’, there are multiple ways attackers can breach a company\u27s container infrastructure and the image supply chain. They also estimated a rise of 600% in the second half of 2020 as compared to the previous year. These attacks generally involves passive scanning methods to access servers that run Docker and Kubernetes platform. Per another report[2], the focus should not only be around securing the tools that are cloud provided, but should also include securing the distributed components involved in the software supply chain throughout the development and deployment process.
This study will lay down a set of rules that can be followed to secure the DevOps workflows for Kubernetes applications, and will cover the most critical security and reliability requirements without causing any delay in the releases and ensure operational independence. This would further detect and prevent attackers who attempt to use Kubernetes to breach the systems, by picking on the vulnerabilities
Online Multiplayer Lobby Creation
No full textA presentation on creating multiplayer lobbies for online games with an interest in user experience, Unity, and Photon
Percentage of Yellow Sour Patch Kids
Get PDFAfter being given the Qualitative Research Project in Introduction to Statistics, I came up with the question asking what percentage of Sour Patch Kids are yellow. This resulted in me going through an entire bag and counting the amount of every color to figure out the percentages. (Class Project