Breast cancer is one of the most common types of cancer worldwide. Early detection and diagnosis are crucial for improving the chances of successful treatment and survival. In this thesis, many different machine learning algorithms were evaluated and compared to predict breast cancer malignancy from diagnostic features extracted from digitized images of breast tissue samples, called fine-needle aspirates. Breast cancer diagnosis typically involves a combination of mammography, ultrasound, and biopsy. However, machine learning algorithms can assist in the detection and diagnosis of breast cancer by analyzing large amounts of data and identifying patterns that may not be discernible to the human eye. By using these algorithms, healthcare professionals can potentially detect breast cancer at an earlier stage, leading to more effective treatment and better patient outcomes. The results showed that the gradient boosting classifier performed the best, achieving an accuracy of 96% on the test set. This indicates that this algorithm can be a useful tool for healthcare professionals in the early detection and diagnosis of breast cancer, potentially leading to improved patient outcomes.

Math homework is a highly debated topic within the middle school education field. Teachers, parents, and students all have differing opinions on what the ideal math homework assignment is and how it promotes academic achievement. This study was intended for discovering what the optimal middle school math homework assignment looks like, how teachers can best follow-up on the assignment, and the most beneficial quantity and frequency of homework. Currently, teachers need more distinct guidelines when designing homework assignments. Students in Barrett, The Honors College, at Arizona State University were asked a series of questions about the type, length, and follow-up practices of their homework assignments and how they felt about them. It was found that students who like math are generally highly motivated in the subject. Most often, students are given short but frequent practice homework assignments, which they find to be most helpful, and they appreciate when teachers review the steps and solutions to the assignments in class. These results should allow educators to better align their math homework assignments with practices that students find to be helpful and necessary.

Many would contend that the United States healthcare system should be moving towards a state of health equity. Here, every individual is not disadvantaged from achieving their true health potential. However, a variety of barriers currently exist that restrict individuals across the country from attaining equitable health outcomes; one of these is the social determinants of health (SDOH). The SDOH are non-medical factors that influence the health outcomes of an individual such as air pollution, food insecurity, and transportation accessibility. Each of these factors can influence the critical illnesses and health outcomes of individuals and, in turn, diminish the level of health equity in affected areas. Further, the SDOH have a strong correlation with lower levels of health outcomes such as life expectancy, physical health, and mental health. Despite having influenced the United States health care system for decades, the industry has only begun to address its influences within the past few years. Through exploration between the associations of the SDOH and health outcomes, programming and policy-making can begin to address the barrier to health equity that the SDOH create.

This paper presents a comprehensive review of current advances and challenges in the field of bone tissue engineering. A systematic review of the literature was conducted to identify recent developments in biomaterials, scaffold design, cell sources, and growth factors for bone tissue engineering applications. Based on this review, an experimental proposal is presented for the development of porous composite biomaterials that may enhance bone regeneration, which consist of hybrid amyloid/spidroin fibers combined with a bioactive ceramic matrix. An iterative design process of modeling and simulation, production, and characterization of both the fibers and the composite material is proposed. A modeling and simulation approach is also presented for unidirectional fiber composite biomaterials using 2-point correlation functions, finite element simulations, and machine learning. This approach was demonstrated to enable the efficient and accurate prediction of the effective Young’s modulus of candidate composite biomaterials, which can inform the design of optimized materials for bone tissue engineering applications. The proposed experimental and simulation approaches have the potential to address current challenges and lead to the development of novel composite biomaterials that can augment the current technologies in the field of bone tissue engineering.

With the increasing popularity of AI and machine learning, human-AI teaming has a wide range of applications in transportation, healthcare, the military, manufacturing, and people’s everyday life. Measurement of human-AI team effectiveness is essential for guiding the design of AI and evaluating human-AI teams. To develop suitable measures of human-AI teamwork effectiveness, we created a search and rescue task environment in Minecraft, in which Artificial Social Intelligence (ASI) agents inferred human teams’ mental states, predicted their actions, and intervened to improve their teamwork (Huang et al., 2022). As a comparison, we also collected data from teams with a human advisor and with no advisor. We investigated the effects of human advisor interventions on team performance. In this study, we examined intervention data and compliance in a human-AI teaming experiment to gain insights into the efficacy of advisor interventions. The analysis categorized the types of interventions provided by a human advisor and the corresponding compliance. The finding of this paper is a preliminary step towards a comprehensive study on ASI agents, in which results from the human advisor study can provide valuable comparisons and insights. Future research will focus on analyzing ASI agents’ interventions to determine their effectiveness, identify the best measurements for human-AI teamwork effectiveness, and facilitate the development of ASI agents.

The main purpose of this project is to create a method for determining the absolute position of an accelerometer. Acceleration and angular speed were obtained from an accelerometer attached to a vehicle as it moves around. As the vehicle moves to collect information the orientation of the accelerometer changes, so a rotation matrix is applied to the data based on the angular change at each time. The angular change and distance are obtained by using the trapezoidal approximation of the integrals. This method was first validated by using simple sets of "true" data which are explicitly known sets of data to compare the results to. Then, an analysis of how different time steps and levels of noise affect the error of the results was performed to determine the optimal time step of 0.1 sec that was then used for the actual tests. The tests that were performed were: a stationary test for uses of calibration, a straight line test to verify a simple test, and a closed loop test to test the accuracy. The graphs for these tests give no indication of the actual paths, so the final results can only show that the data from the accelerometer is too noisy and inaccurate for this method to be used by this sensor. The future work would be to test different ways to get more accurate data and then use it to verify this methods. These ways could include using more sensors to interpolate the data, reducing noise by using a different sensor, or adding a filter. Then, if this method is considered accurate enough, it could be implemented into control systems.

The incidence of childhood obesity has become increasingly prevalent in the United States in recent years. The development of obesity at any age, but especially in adolescence, can have lasting negative effects in the form of cardiometabolic disease, increased incurred healthcare costs, and potential negative effects on quality of life. In recent years, a rising trend of obesity, in both adults and adolescents, has been observed in lower income and ethnic groups. Increased adiposity can be influenced by modifiable factors -(physical activity, caloric intake, or sleep) or by non-modifiable factors (ethnicity, genetic predispositions, and socioeconomic status). The influence of these factors can be observed in individuals of all ages, including infants. A common indicator of the development of childhood obesity is rapid weight gain (RWG) within an infant’s first year of life. The composition of the gut microbiome can act as a predictor for RWG and the development of childhood obesity. Infants are exposed to an immense microbial load when they are born and their gut microbiome is continually diversified through their method of feeding and the subsequent introduction to solid foods. While currently understudied, it is understood that cultural and socioeconomic factors influence the development of the gut microbiome, which is further explored in this analysis. The DNA from 51 fecal samples from infants ranging from 3 weeks to 12 months in age was extracted and sequenced using next-generation sequencing, and the resulting sequences were analyzed using QIIME 2. Results from alpha-diversity and beta-diversity metrics showed significant differences in the gut microbiome of infants when comparing groups based on baby race/ethnicity, household income, and mom’s education. These findings suggest the importance of sociodemographic characteristics in shaping the gut microbiome and suggest the importance of future studies including diverse populations in gut microbiome work.

This paper serves as an analysis of the current operational conditions of a real-world company – referred to as “Company X” – with respect to the IC substrate industry. The cost of substrates, a crucial component in the production of Company X’s product, has recently diverged from Company X’s predictions and is contributing to declining profitability. This analysis aims to discover the underlying cause for price divergence and recommend potential resolutions to improve the forecast of substrate costs and profitability. The paper is organized as follows: Chapter 1 is an introduction to IC substrates and the industry as a whole, Chapter 2 is a breakdown of the specific factors responsible for substrate prices, and Chapter 3 delivers a final recommendation to Company X and concludes the paper.

Graph neural networks (GNN) offer a potential method of bypassing the Kohn-Sham equations in density functional theory (DFT) calculations by learning both the Hohenberg-Kohn (HK) mapping of electron density to energy, allowing for calculations of much larger atomic systems and time scales and enabling large-scale MD simulations with DFT-level accuracy. In this work, we investigate the feasibility of GNNs to learn the HK map from the external potential approximated as Gaussians to the electron density 𝑛(𝑟), and the mapping from 𝑛(𝑟) to the energy density 𝑒(𝑟) using Pytorch Geometric. We develop a graph representation for densities on radial grid points and determine that a k-nearest neighbor algorithm for determining node connections is an effective approach compared to a distance cutoff model, having an average graph size of 6.31 MB and 32.0 MB for datasets with 𝑘 = 10 and 𝑘 = 50 respectively. Furthermore, we develop two GNNs in Pytorch Geometric, and demonstrate a decrease in training losses for a 𝑛(𝑟) to 𝑒(𝑟) of 8.52 · 10^14 and 3.10 · 10^14 for 𝑘 = 10 and 𝑘 = 20 datasets respectively, suggesting the model could be further trained and optimized to learn the electron density to energy functional.

This thesis focuses on how domain formation and local disorder mediate non-equilibrium order in the context of condensed matter physics. More specifically, the data supports c-axis CDW ordering in the context of the rare-earth Tritellurides. Experimental studies were performed on Pd:ErTe3 by ultra-fast pump-probe and x-ray free electron laser (XFEL). Ginzburg Landau models were used to simulate domain formation. Universal scaling analysis on the data reveals that topological defects govern the relaxation of domain walls in Pd:ErTe3. This thesis presents information on progress towards using light to control material domains.