The goal of this project is to explore how 23 different bird species respond to 3 climatic attributes. These attributes are lower than average temperatures, wind speed and precipitation level. Information about the bird species and all of the data associated with them is provided by Project FeederWatch (PFW). This is a citizen based survey study that provides key information about bird species abundance through the use of backyard and community feeders. The study volunteers from across the United States and Canada monitor these bird feeders and note important information about the species such as the number of individuals seen. Other standard information is also included such as location data and date. An original data collection pipeline was developed for this study to append climate data from Weather Underground (WU) to the PFW bird feeder data. The final dataset helped to explore how exactly the birds are reacting to winter temperatures, wind speeds and rain levels. Our results indicate that birds species in general visit the bird feeders more often as temperatures dip below average. We found that the body mass of the bird plays no role in the number of visits. Birds don't seem to be significantly affected by precipitation or wind speed as our results indicate no relationship between these climatic factors and abundance at the feeders.

Several regions of the Western United States utilize statistical binary classification models to predict and manage debris flow initiation probability after wildfires. As the occurrence of wildfires and large intensity rainfall events increase, so has the frequency in which development occurs in the steep and mountainous terrain where these events arise. This resulting intersection brings with it an increasing need to derive improved results from existing models, or develop new models, to reduce the economic and human impacts that debris flows may bring. Any development or change to these models could also theoretically increase the ease of collection, processing, and implementation into new areas. Generally, existing models rely on inputs as a function of rainfall intensity, fire effects, terrain type, and surface characteristics. However, no variable in these models directly accounts for the shear stiffness of the soil. This property when considered with the respect to the state of the loading of the sediment informs the likelihood of particle dislocation, contractive or dilative volume changes, and downslope movement that triggers debris flows. This study proposes incorporating shear wave velocity (in the form of slope-based thirty-meter shear wave velocity, Vs30) to account for this shear stiffness. As commonly used in seismic soil liquefaction analysis, the shear stiffness is measured via shear wave velocity which is the speed of the vertically propagating horizontal shear wave through sediment. This spatially mapped variable allows for broad coverage in the watersheds of interest. A logistic regression is used to then compare the new variable against what is currently used in predictive post-fire debris flow triggering models. Resulting models indicated improvement in some measures of statistical utility through receiver operating characteristic curves (ROC) and threat score analysis, a method of ranking models based on true/false positive and negative results. However, the integration of Vs30 offers similar utility to current models in additional metrics, suggesting that this input can benefit from further refinement. Further suggestions are additionally offered to further improve the use of Vs30 through in-situ measurements of surface shear wave propagation and integration into Vs30 datasets through a possible transfer function. Additional discussion into input variables and their impact on created models is also included.

Pervasive growth in industrialization and advances in technology now exposes much of the world to anthropogenic night light and noise (ANLN), which pose a global environmental challenge in terrestrial environments. An estimated one-tenth of the planet’s land area experiences artificial light at night — and that rises to 23% if skyglow is included. Moreover, anthropogenic noise is associated with urban development and transportation networks, as the ecological impact of roads alone is estimated to affect one-fifth of the total land cover of the United States and is increasing in space and intensity. Existing research involving impacts of light or noise has primarily focused on a single sensory stressor and single species; yet, little information is known about how different sources of sensory stressors impact the relationships within tightly-knit and complex systems, such as within plant-pollinator communities. Furthermore, ANLN often co-occur, yet little is known about how co-exposure to these stressors influences wildlife, nor the extent and scale of how these stressors impact ecological processes and patterns. In Chapter 1, we had two aims: to investigate species-specific responses to artificial night light, anthropogenic noise, and the interaction between the two by using spatially-explicit models to model changes in abundance of 140 of the most prevalent overwintering bird species across North America, and to identify functional traits and contexts that explain variation in species-specific responses to ANLN stressors with phylogenetically-informed models. We found species that responded to noise exposure generally decreased in abundance, and the interaction with light resulted in negative synergistic responses that exacerbated the negative influence of noise among many species. Moreover, the interaction revealed negative emergent responses of species that only reacted when both ANLN were presented in combination. The functional trait that was the most indicative of avian response to ANLN was habitat preference. Specifically, species that occupy closed habitat were less tolerant of both sensory stressors compared to those that occupy open habitat. Species-specific responses to ANLN are context-dependent; thus, knowing the information that regulates when, where, how, and why sensory pollutants influence species will help management efforts effectively mitigate these anthropogenic stressors on the natural environment. In Chapter 2, using field-placed light manipulations at sites exposed to a gradient of skyglow, we investigated the influence of direct and indirect light on the yucca-yucca moth mutualism by quantifying chaparral yucca (Hesperoyucca whipplei) fruit set and the obligate moth (Tegeticula maculata maculata) larval density per fruit. Although many diurnal insects are thought to exhibit minimal phototaxis, we show that direct light attracted adult moths and incited higher pollination activity, resulting in an increase in fruit set. However, larval recruitment decreased with elevated light exposure and the effect was strongest for plants exposed to light levels exceeding natural moonlit conditions (> 0.5 lux). Contrarily, increases in ambient skyglow resulted in an increase in both fruit set and larva counts. Our results suggest that plant-pollinator communities may respond in complicated ways to different sources of light, such that novel selection pressures of direct and indirect light have the potential to benefit or disrupt networks within complex diurnal plant-pollinator communities, and ultimately alter the biodiversity reliant on these systems. By analyzing pervasive stressors across a continental-wide scale, we revealed considerable heterogeneity in avian responses to light and noise alone, as well as the interaction between them. Based on overall responses to the interaction between light v and noise, we suggest management efforts should focus on ameliorating excessive noise for overwintering bird species, which should decrease the impact from synergistic responses, as well as the negative impact from noise alone. There is still much to learn about responses to these stressors and smaller-scale studies should take our approach of systematically assessing interaction responses to ANLN. Moreover, our small-scale study revealed both local sources of direct light and skyglow impact the recruitment for both yucca moths and their reciprocal plant hosts. However, it is still unknown if or why other diurnal pollinators experience positive phototaxis, and whether direct lighting influences the physiology, behavior, or multiple factors relating to reproduction and fitness. Correspondingly, it is unknown if the novel selection pressures of direct and indirect light are disrupting complex diurnal plant-pollinator communities. Future research on artificial night light will need to investigate the intricate responses of diurnal pollinators to both direct and indirect light that will identify concrete mechanisms relating to physiological or behavioral susceptibility and inform predictions on how wide-spread communities will shift with this global driver of emerging change.

Infectious disease forecasting efforts underwent rapid growth during the COVID-19 pandemic, providing guidance for pandemic response and about potential future trends. Yet despite their importance, short-term forecasting models often struggled to produce accurate real-time predictions of this complex and rapidly changing system. This gap in accuracy persisted into the pandemic and warrants the exploration and testing of new methods to glean fresh insights. In this work, we examined the application of the temporal hierarchical forecasting (THieF) methodology to probabilistic forecasts of COVID-19 incident hospital admissions in the United States. THieF is an innovative forecasting technique that aggregates time-series data into a hierarchy made up of different temporal scales, produces forecasts at each level of the hierarchy, then reconciles those forecasts using optimized weighted forecast combination. While THieF's unique approach has shown substantial accuracy improvements in a diverse range of applications, such as operations management and emergency room admission predictions, this technique had not previously been applied to outbreak forecasting. We generated candidate models formulated using the THieF methodology, which differed by their hierarchy schemes and data transformations, and ensembles of the THieF models, computed as a mean of predictive quantiles. The models were evaluated using weighted interval score (WIS) as a measure of forecast skill, and the top-performing subset was compared to several benchmark models. These models included simple ARIMA and seasonal ARIMA models, a naive baseline model, and an ensemble of operational incident hospitalization models from the US COVID-19 Forecast Hub. The THieF models and THieF ensembles demonstrated improvements in WIS and MAE, as well as competitive prediction interval coverage, over many benchmark models for both the validation and testing phases. The best THieF model generally ranked second out of nine total models during the testing evaluation. These accuracy improvements suggest the THieF methodology may serve as a useful addition to the infectious disease forecasting toolkit.

The increasing demand for healthcare services, coupled with the challenges of managing budgets and navigating complex regulations, has underscored the need for sustainable and efficient healthcare delivery. In response to this pressing issue, this thesis aims to optimize hospital efficiency using Artificial Intelligence (AI) techniques. The focus extends beyond improving surgical intraoperative time to encompass preoperative and postoperative periods as well. The research presents a novel Prescriptive Analytics System (PAS) designed to enhance the Surgical Success Rate (SSR) in surgeries and specifically in high volume arthroplasty. The SSR is a critical metric that reflects the successful completion of 4-surgeries during an 8-hour timeframe. By leveraging AI, the developed PAS has the potential to significantly improve the SSR from its current rate of 39% at The Ottawa Hospital to a remarkable 100%. The research is structured around five peer-reviewed journal papers, each addressing a specific aspect of the optimization of surgical efficiency. The first paper employs descriptive analytics to examine the factors influencing delays and overtime pay during surgeries. By identifying and analyzing these factors, insights are gained into the underlying causes of surgery inefficiencies. The second paper proposes three frameworks aimed at improving Operating Room (OR) throughput. These frameworks provide structured guidelines and strategies to enhance the overall efficiency of surgeries, encompassing preoperative, intraoperative, and postoperative stages. By streamlining the workflow and minimizing bottlenecks, the proposed frameworks have the potential to significantly optimize surgical operations. The third paper outlines a set of actions required to transform a selected predictive system into a prescriptive one. By integrating AI algorithms with decision support mechanisms, the system can offer actionable recommendations to surgeons during surgeries. This transformative step holds tremendous potential in enhancing surgical outcomes while reducing time. The fourth paper introduces a benchmarking and monitoring system for the selected framework that predicts SSR. Leveraging historical data, this system utilizes supervised machine learning algorithms to forecast the likelihood of successful outcomes based on various surgical team and procedural parameters. By providing real-time monitoring and predictive insights, surgeons can proactively address potential risks and improve decision-making during surgeries. Lastly, an application paper demonstrates the practical implementation of the prescriptive analytics system. The case study highlights how the system optimizes the allocation of resources and enables the scheduling of additional surgeries on days with a high predicted SSR. By leveraging the system's capabilities, hospitals can maximize their surgical capacity and improve overall patient care.

Recently, strategies of National Basketball Association teams have evolved with the skillsets of players and the emergence of advanced analytics. One of the most effective actions in dynamic offensive strategies in basketball is the dribble hand-off (DHO). This thesis proposes an architecture for a classification pipeline for detecting DHOs in an accurate and automated manner. This pipeline consists of a combination of player tracking data and event labels, a rule set to identify candidate actions, manually reviewing game recordings to label the candidates, and embedding player trajectories into hexbin cell paths before passing the completed training set to the classification models. This resulting training set is examined using the information gain from extracted and engineered features and the effectiveness of various machine learning algorithms. Finally, we provide a comprehensive accuracy evaluation of the classification models to compare various machine learning algorithms and highlight their subtle differences in this problem domain.

<p dir="ltr">Lithium-ion batteries are being used increasingly more often to provide power for systems that range all the way from common cell-phones and laptops to advanced electric automotive and aircraft vehicles. However, as is the case for all battery types, lithium-ion batteries are prone to naturally occurring degradation phenomenon that limit their effective use in these systems to a finite amount of time. This degradation is caused by a plethora of variables and conditions including things like environmental conditions, physical stress/strain on the body of the battery cell, and charge/discharge parameters and cycling. Accurately and reliably being able to predict this degradation behavior in battery systems is crucial for any party looking to implement and use battery powered systems. However, due to the complicated non-linear multivariable processes that affect battery degradation, this can be difficult to achieve. Compared to traditional methods of battery degradation prediction and modeling like equivalent circuit models and physics-based electrochemical models, data-driven machine learning tools have been shown to be able to handle predicting and classifying the complex nature of battery degradation without requiring any prior knowledge of the physical systems they are describing.</p><p dir="ltr">One of the most critical steps in developing these data-driven neural network algorithms is data procurement and preprocessing. Without large amounts of high-quality data, no matter how advanced and accurate the architecture is designed, the neural network prediction tool will not be as effective as one trained on high quality, vast quantities of data. This work aims to gather battery degradation data from a wide variety of sources and studies, examine how the data was produced, test the effectiveness of the data in the Interfacial Multiphysics Laboratory’s autoencoder based neural network tool CD-Net, and analyze the results to determine factors that make battery degradation datasets perform better for use in machine learning/deep learning tools. This work also aims to relate this work to other data-driven models by comparing the CD-Net model’s performance with the publicly available BEEP’s (Battery Evaluation and Early Prediction) ElasticNet model. The reported accuracy and prediction models from the CD-Net and ElasticNet tools demonstrate that larger datasets with actively selected training/testing designations and less errors in the data produce much higher quality neural networks that are much more reliable in estimating the state-of-health of lithium-ion battery systems. The results also demonstrate that data-driven models are much less effective when trained using data from multiple different cell chemistries, form factors, and cycling conditions compared to more congruent datasets when attempting to create a generalized prediction model applicable to multiple forms of battery cells and applications.</p>

The availability of a scalable and explainable rule extraction technique via motif discovery is crucial for identifying the health states of a system. Such a technique can enable the creation of a repository of normal and abnormal states of the system and identify the system’s state as we receive data. In complex systems such as ECG, each activity session can consist of a long sequence of motifs that form different global structures. As a result, applying machine learning algorithms without first identifying the local patterns is not feasible and would result in low performance. Thus, extracting unique local motifs and establishing a database of prototypes or signatures is a crucial first step in analyzing long temporal data that reduces the computational cost and overcomes imbalanced data. The present research aims to streamline the extraction of motifs and add explainability to their analysis by identifying their differences. We have developed a novel framework for unsupervised motif extraction. We also offer a robust algorithm to identify unique motifs and their signatures, coupled with a proper distance metric to compare the signatures of partially similar motifs. Defining such distance metrics allows us to assign a degree of semblance between two motifs that may have different lengths or contain noise. We have tested our framework against five different datasets and observed excellent results, including extraction of motifs from 100 million samples in 8.02 seconds, 99.90% accuracy in self-supervised ECG data classification, and an average error of 16.66% in RUL prediction of bearing failure.

<p dir="ltr">One of the issues on social media that is expanding the fastest is children being exposed to predators online [ 1 ]. Due to the ease with which a larger segment of the younger population may now access the Internet, online grooming activity on social media has grown to be a significant social concern. Child grooming, in which adults and minors exchange sexually explicit text and media via social media platforms, is a typical component of online child exploitation. An estimated 500,000 predators operate online every day. According to estimates, Internet chat rooms and instant messaging are where 89% of sexual approaches against children take place. The child may face a variety of unpleasant consequences following a grooming event, including shame, anger, anxiety, tension, despair, and substance abuse which make it more difficult for them to report the exploitation. A substantial amount of research in this domain has focused on identifying certain vulnerabilities of the victims of grooming. These vulnerabilities include specific age groups, gender, psychological factors, no family support, and lack of good social relations which make young people more vulnerable to grooming. So far no technical work has been done to apply statistical analysis on these vulnerability profiles and observe how these patterns change between different victim types and offender types. This work presents a detailed analysis of the effect of Offender type (contact and fantasy) and victim type (Law Enforcement Officers, Real Victims and Decoys (Perverted Justice)) on representation of different vulnerabilities in grooming conversations. Comparison of different victim groups would provide insights into creating the right training material for LEOs and decoys and help in the training process for online sting operations. Moreover, comparison of different offender types would help create targeted prevention strategies to tackle online child grooming and help the victims.</p>

Historical manuscripts are an essential source of original content. For many reasons, it is hard to recognize these manuscripts as text. This thesis used a state-of-the-art Handwritten Text Recognizer, TrOCR, to recognize a 16th-century manuscript. TrOCR uses a vision transformer to encode the input images and a language transformer to decode them back to text. We showed that carefully preprocessed images and designed augmentations can improve the performance of TrOCR. We suggest an ensemble of augmented models to achieve an even better performance.