Filtering by
- Member of: Barrett, The Honors College Thesis/Creative Project Collection
- Member of: ASU Electronic Theses and Dissertations
Covid-19 is unlike any coronavirus we have seen before, characterized mostly by the ease with which it spreads. This analysis utilizes an SEIR model built to accommodate various populations to understand how different testing and infection rates may affect hospitalization and death. This analysis finds that infection rates have a significant impact on Covid-19 impact regardless of the population whereas the impact that testing rates have in this simulation is not as pronounced. Thus, policy-makers should focus on decreasing infection rates through targeted lockdowns and vaccine rollout to contain the virus, and decrease its spread.
An examination upon the historical evolution of the quarterback reveals that there were three foundational cycles leading up to 2007 which established the model for the mobile quarterback in the NFL. These were especially marked by exceptional quarterbacks breaking molds and pioneering African American quarterbacks overcoming racial stigma. Since 2007, there has been a steady trend of mobile quarterbacks replacing pocket passers, especially among playoff teams. Using k-means clustering, three different categories of quarterbacks were established: pocket passers, scramblers, and dual-threats. After evaluating various player metrics describing quarterback mobility, using yards per game, run-to-pass ratio, scramble rate, and designed run rate on third down produced the best model. This yielded an accurate prediction of covariance and a good overall fit. Teams with dual-threat quarterbacks had more success than other quarterback types on third-and-medium for dropbacks, third-and-long for designed runs, and explosive plays (plays which gain 20+ yards) on designed runs, passes, and quarterback scrambles. An examination into the schematic tendencies using film reveals that mobile quarterbacks allow the offense to have more freedom in its play calling and reduces the margin of error for defenses. Alongside the NFL’s increased focus on the concept of positionless football, this provides the framework for what this thesis calls the “Slashback Offense,” in which the offense utilizes a young, athletic quarterback in multiple positions in conjunction with a mobile starting quarterback. This can enhance option plays, establish the threat of another passer, and reduce the physical burden on the starting quarterback.
Under sparsity, the structure of DSDs can allow for the screening and optimization of a system in one step, but in non-sparse situations estimation of second-order models requires augmentation of the DSD. In this work, augmentation strategies for DSDs were considered, given the assumption that the correct form of the model for the response of interest is quadratic. Series of augmented designs were constructed and explored, and power calculations, model-robustness criteria, model-discrimination criteria, and simulation study results were used to identify the number of augmented runs necessary for (1) effectively identifying active model effects, and (2) precisely predicting a response of interest. When the goal is identification of active effects, it is shown that supersaturated designs are sufficient; when the goal is prediction, it is shown that little is gained by augmenting beyond the design that is saturated for the full quadratic model. Surprisingly, augmentation strategies based on the I-optimality criterion do not lead to better predictions than strategies based on the D-optimality criterion.
Computational limitations can render standard statistical methods infeasible in the face of massive datasets, necessitating subsampling strategies. In the big data context, the primary objective is often prediction but the correct form of the model for the response of interest is likely unknown. Here, two new methods of subdata selection were proposed. The first is based on clustering, the second is based on space-filling designs, and both are free from model assumptions. The performance of the proposed methods was explored visually via low-dimensional simulated examples; via real data applications; and via large simulation studies. In all cases the proposed methods were compared to existing, widely used subdata selection methods. The conditions under which the proposed methods provide advantages over standard subdata selection strategies were identified.
