Physical Sciences and Mathematics Commons^™

Neural Network Pruning For Ecg Arrhythmia Classification, Isaac E. Labarge

Master's Theses

Convolutional Neural Networks (CNNs) are a widely accepted means of solving complex classification and detection problems in imaging and speech. However, problem complexity often leads to considerable increases in computation and parameter storage costs. Many successful attempts have been made in effectively reducing these overheads by pruning and compressing large CNNs with only a slight decline in model accuracy. In this study, two pruning methods are implemented and compared on the CIFAR-10 database and an ECG arrhythmia classification task. Each pruning method employs a pruning phase interleaved with a finetuning phase. It is shown that when performing the scale-factor pruning …

Seasonal And Interannual Variability Of Phytoplankton Abundance And Community Composition On The Central Coast Of California, Alex Barth, Ryan K. Walter, Ian C. Robbins, Alexis Pasulka

Physics

Variations in the abundance and composition of phytoplankton greatly impact ecosystem structure and function. Within the California Current System (CCS), phytoplankton community structure is tightly coupled to seasonal variability in wind-driven coastal upwelling, a process that drives changes in coastal water temperatures and nutrient concentrations. Based on approximately a decade (2008-2018) of weekly phytoplankton measurements, this study provides the first characterization of the seasonal and interannual variability of phytoplankton abundance and composition in San Luis Obispo (SLO) Bay, an understudied region within the CCS. Overall, the seasonality of phytoplankton in SLO Bay mirrored that of the larger CCS; diatoms dominated …

Valorization Of Carrot Processing Waste, Alexandra M. Duval

Master's Theses

Commercial carrot processors produce up to 175,000 tons of carrot waste annually. Carrot Mash (CM) is the term referring to the waste by-product of peeled baby carrot processing. Transportation of carrot processing waste is expensive due to its high- water content (approx. 83-95%). High in bioactive compounds (carotenoids) and dietary fibers, it is expected that its conversion into a value-added by-product is of interest to the carrot processing industry. Hemicellulose-rich plant materials have proven to be a source of oligosaccharides, which are known for their beneficial prebiotic activity. The objectives of this research were to: 1) determine the effect of …

Optimizing Llrf Parameters In The Electron-Ion Collider, William M. Bjorndahl

Physics

To improve particle interaction in the future Electron-Ion Collider (EIC), we investigated different feedback implementations to control the accelerating voltage and examined the power and beam phase for each instance. Using MATLAB, we studied three feedback mechanisms: Direct, One Turn, and Feedforward. Enacting feedforward yielded the best performance. To minimize the klystron power consumption, we analyzed different Low-Level Radio Frequency (LLRF) parameters such as detuning. Combining theory and simulated results, we found the optimal detuning value that minimizes klystron power consumption.

Host-Galaxy Morphology Of 14 Local Active Galaxies As Imaged By The Hubble Space Telescope, Tomas Henry Snyder

Physics

Active galactic nuclei (AGNs) are galaxies with super-massive black holes, that are undergoing accretion, at their centers. The study of these objects provides us with a deeper understanding of how the black holes evolved along with their host galaxies since black hole mass is highly correlated with bulge mass. A sample of 66 AGNs were imaged using the Hubble Space Telescope (HST) and 14 of those are analyzed in this paper to determine their structure, bulge luminosity, and bulge type. The program GALFIT is implemented to determine these parameters, that will be used in further research in the future. These …

Llama: The MBh–Σ⋆ Relation Of The Most Luminous Local Agns, Turgay Caglar, Leonard Burtscher, Bernhard Brandl, Jarle Brinchmann, Richard I. Davies, Erin K. S. Hicks, Michael Koss, Ming-Yi Lin, Witold Maciejewski, Francisco MüLler-Sánchez, Rogemar A. Riffel, Rogério Riffel, David J. Rosario, Marc Schartmann, Allan Schnorr-MüLler, T. Taro Shimizu, Thaisa Storchi-Bergmann, Sylvain Veilleux, Gilles Orban De Xivry, Vardha N. Bennert

Physics

Context. The MBH–σ_⋆ relation is considered a result of coevolution between the host galaxies and their supermassive black holes. For elliptical bulge hosting inactive galaxies, this relation is well established, but there is still discussion concerning whether active galaxies follow the same relation.

Aims. In this paper, we estimate black hole masses for a sample of 19 local luminous active galactic nuclei (AGNs; LLAMA) to test their location on the MBH–σ_⋆ relation. In addition, we test how robustly we can determine the stellar velocity dispersion in the presence of an AGN continuum and AGN emission lines, …

Automating Deep-Sea Video Annotation Using Machine Learning, Hanson Egbert, Lubomir Stanchev, Benjamin Ruttenberg

Computer Science and Software Engineering

As the world explores opportunities to develop offshore renewable energy capacity, there will be a growing need for pre-construction biological surveys and post-construction monitoring in the challenging marine environment. Underwater video is a powerful tool to facilitate such surveys, but the interpretation of the imagery is costly and time-consuming. Emerging technologies have improved automated analysis of underwater video, but these technologies are not yet accurate or accessible enough for widespread adoption in the scientific community or industries that might benefit from these tools. To address these challenges, we developed a website that allows us to: (1) Quickly play and annotate …

Hydrologic Response Of A Montane Meadow From Conifer Removal And Upslope Forest Thinning, Christopher Surfleet, Noel Fie, John Jasbinsek

Physics

This study evaluates the hydrologic response of restoration of a montane meadow by removal of encroached Pinus contorta and thinning of the adjacent forest. It is now a follow-up with four years of post-restoration data, on a previous analysis of a hydrologic response of the same meadow one year following restoration. A hydrologic change was evaluated through a statistical comparison of soil moisture and depth to groundwater between the restored Marian Meadow and a Control Meadow. Meadow water budgets and durations of water table depths during the growing season were evaluated. The four years following restoration of Marian Meadow had …

The Mass Relations Between Supermassive Black Holes And Their Host Galaxies At 1 < Z < 2 With Hst-Wfc3, Xuheng Ding, John Silverman, Tommaso Treu, Andreas Schulze, Malte Schramm, Simon Birrer, Daeseong Park, Knud Jahnke, Vardha N. Bennert, Jeyhan S. Kartaltepe, Anton M. Koekemoer, Matthew A. Malkan, David Sanders

Physics

Correlations between the mass of a supermassive black hole (SMBH) and the properties of its host galaxy (e.g., total stellar mass M*, luminosity Lhost) suggest an evolutionary connection. A powerful test of a coevolution scenario is to measure the relations –Lhost and –M* at high redshift and compare with local estimates. For this purpose, we acquired Hubble Space Telescope (HST) imaging with WFC3 of 32 X-ray-selected broad-line (type 1) active galactic nuclei at 1.2 < z < 1.7 in deep survey fields. By applying state-of-the-art tools to decompose the HST images including available ACS data, we measured the host galaxy luminosity and stellar mass along with …

Human Path Prediction Using Auto Encoder Lstms And Single Temporal Encoders, Hayden Hudgins

Master's Theses

Due to automation, the world is changing at a rapid pace. Autonomous agents have become more common over the last several years and, as a result, have created a need for improved software to back them up. The most important aspect of this greater software is path prediction, as robots need to be able to decide where to move in the future. In order to accomplish this, a robot must know how to avoid humans, putting frame prediction at the core of many modern day solutions. A popular way to solve this complex problem of frame prediction is Auto Encoder …

Intelligent Cinematic Camera Control For Real-Time Graphics Applications, Ian Harris Meeder

Master's Theses

E-sports is currently estimated to be a billion dollar industry which is only growing in size from year to year. However the cinematography of spectated games leaves much to be desired. In most cases, the spectator either gets to control their own freely-moving camera or they get to see the view that a specific player sees. This thesis presents a system for the generation of cinematically-pleasing views for spectating real-time graphics applications. A custom real-time engine has been built to demonstrate the effect of this system on several different game modes with varying visual cinematic constraints, such as the rule …