18736 research outputs found
Sort by
Application for Prediction of Heart Failure; the Next Step in Machine Learning for Healthcare
Get PDFHeart failure (HF) is a serious medical condition affecting approximately 6.7 million U.S. adults and is expected to impact 8.5 million Americans by 2030 [1]. Heart failure is a complicated clinical ailment and characterizes the final course of numerous heart diseases [2]. This paper introduces a machine-learning-based application that utilizes Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), and XGBoost models, implemented through the Python Flask framework, to predict HF risk using clinical data. The results indicate high model performance, with precision and recall metrics underscoring the application’s reliability in identifying at-risk patients. By providing real-time, accessible insights, this tool aims to enhance diagnostic accuracy and support early intervention, withparticular value for underserved populations. The study also discusses limitations related to data access, privacy concerns, and model generalizability, recommending future research to address these challenges for broader clinical adoption
Scaling Survivor Restoration through Data Science
Get PDFInternational Justice Mission (IJM) has identified and service over 50,000 survivors of human trafficking around the world. This data-informed approach uses Assessment of Survivor Outcomes (ASO) captured in the case-tracking system called Justice Data Solution to support Aftercare managers around the world in their care of survivors. The approach centers on critical, time-sensitive needs for survivors at the point of rescue, at the mid-point of their restoration journey, and at end of their time with our Aftercare program. During this program, survivors are engaged in community events, health services, and legal proceedings. Using the data from the ASO and the services/events, insights can be draw on the Aftercare program and where more value can be created for survivors to allow them to quickly be restored in their community. The analysis reveals the optimal time for survivors to be in the Aftercare program, the specific indicators in the initial ASO that highlight survivors requiring more support, and how we might optimize delivery of services and train partner organizations on the same. This analysis allows local IJM teams to scale efforts to restore survivors and empower local survivor leaders in the fight for justice
Best Practices: Reusing Space to Enhance Narrative Immersion
Get PDFThis research explores the relationship between traditional storytelling techniques and the reuse of space and attempts to evaluate the relationship’s effectiveness in enhancing the player’s immersion in both narrative and gameplay. The researcher investigates various concepts, such as mental mapping, spaced learning, and thematic affordances, to understand how players remember previously visited spaces. Based on these practices, the researcher built a custom single-player level in Dying Light, where players assist a trapped non-player character (NPC) in searching for his coworkers. As players complete objectives, they simultaneously familiarize themselves with the environment and revisit spaces guided by narrative signifiers. To assess the impact of spatial reuse on a player’s narrative and gameplay experience, the researcher collected data from multiple playtests and analyzed participants\u27 comprehension of the story and level layout
Predicting Subversion Impact in a Narratively Driven Quest
Get PDFThis thesis aimed to develop a methodology which can be used to quantify a narrative subversion’s memorability. The researcher formulated seven unique subversion categories: Action, Environment, Narrative, Betrayal, Genre, Stereotype and Pattern. The researcher created a single player video game level in Fallout 4. The researcher developed a quest incorporating seven subversions (two Action, two Environment, two Narrative, and one Betrayal). To evaluate each subversion, the researcher created a “subversion memorability” formula using four key variables (Impact, Severity, Time, and Number of Prior Subversions). The researcher then plotted the subversion memorability values on a line graph to examine whether the resulting curve aligned with the Three-Act Structure. The researcher then conducted playtests and asked participants to indicate the levels of impact and severity of each subversion. The researcher then plugged the response averages into the formula to calculate the true subversion memorability value of each individual subversion
Exploring the Relationship Between Architectural Accuracy and Gameplay Flow and Its Impact on Player Immersion
Get PDFThis thesis topic discusses the relationship between architectural accuracy and gameplay flow and the impact it has on the player’s immersion. The researcher created two levels, a control and a test level, to compare a level that is focused primarily on architectural accuracy and another level that breaks down or “warps” the architectural accuracy in favor of improving gameplay flow. The researcher aimed to explore the important aspects of level building, specifically the aesthetics, architectural accuracy, and gameplay in each space and how those qualities affect the player’s immersion
Best Practices: A Framework for Modulating Melee and Ranged Combat Styles
Get PDFThis thesis investigates the impact of video game environmental design, resource abundance, and fault tolerance on players\u27 in-game combat styles between melee and ranged. The researcher integrates all three aspects into a single-player level, Dam Escape, built using Dying Light Developer Tools, to collect quantitative and qualitative player behavior data. By aligning predicted outcomes with actual gameplay data, this research seeks to refine actionable insights for balancing player agency with intentional level design. The results, derived from regression analysis and iterative calibration, will be presented in the data and conclusions
An Exploration into Liminal Spaces and Their Effects on Player Tension
Get PDFThis thesis explores how liminal spaces can be used in video games by analyzing their design and impact on player tension. How can designers create transitional or ambiguous environments that contribute to a sense of unease, anticipation, and tension? What are some of the common design elements that help build an effective liminal space, which the researcher calls distortions, and how do those distortions affect player tension? The researcher created a series of levels that explore the different elements of liminal spaces to test how they affect player tension while playing the level
Using Object Density to Modulate Tension Responses
Get PDFThis thesis investigates the connection between object density and player tension in video games. Expanding on Canny Yuan’s work, “Using Spatial Composition to Influence Player Tension”, the researcher introduced a quantitative approach to measuring room density and posits that increased navigable space can lower player tension. To test this hypothesis, the researcher developed a custom single-player level, “Veylmoor”, in The Elder Scrolls V: Skyrim. The level is composed of rooms with systematically varied spatial layouts and navigable areas, allowing for controlled analysis of spatial density. The researcher quantified room density using four methodologies implemented through custom C# scripts. The researcher also assessed player tension through heart rate data and a post-playtest survey, enabling a comprehensive evaluation of the proposed methodologies
Advanced Mathematical and Computational Methods in Quantum Chemistry
Get PDFWave-function based ab initio calculation methods have been gaining significant atten- tion from chemists due to their ability to provide insights into experimental results. However, their application to large systems is limited by the considerable computational cost involved, especially when dealing with high-dimensional tensors inherent to these calculations. To address these challenges, two primary strategies have been explored in this study. First, advancements in hardware architecture and algorithmic development. Graphics Processing Units (GPUs) have facilitated parallel computation, leading to a significant acceleration of these calculations. In this work, we propose a method for evaluating two-electron repulsion integrals (ERIs)—a key component of quantum chemistry—on GPUs. Regarding to the algorithm development, we propose a series of new algorithms for skew symmetric matrix LTLT decomposition,andimplementedhighperformanceCPUimplementationtoaccelerate these matrix operations in Quantum Monte Carlo (QMC) simulation./= / \u3eSecond, the development of approximation methods, such as tensor hypercontraction (THC), specifically least-squares tensor (LS-THC), has been introduced to reduce the high- dimensional tensors commonly encountered in computational chemistry, which allows wave- function methods to be applied to larger systems. However, the accuracy of LS-THC depends on the grid points. In this work, we proposed a pair of grid point schemes to account for missing interactions for grid points. In addition, we propose a novel energy pivoting algorithm to select grid points to prune unnecessary ones, which demonstrates good performance over the traditional pruning algorithm