No abstract

Abstract-This study explores the effect of incorporating an Open-Ended Design Experience (OEDE) into an undergraduate materials science laboratory taken by third-year mechanical engineering students. The focus of the OEDE was carbon fiber reinforced plastics and sandwich structures. The results indicate that the incorporation of OEDEs in laboratory courses produces significant benefits in terms of student engagement, participation, and perception of competence. In addition, the OEDE was found to enhance students' ability to apply related concepts as compared to non-OEDE lab activities. The authors conclude that the incorporation of OEDEs can increase the effectiveness of engineering laboratory courses.

No abstract

Annual salmon migrations vary significantly in annual return numbers from year to year. In order to determine when a species' sustainable return size has been met, a method for counting and sizing the spawning animals is required. This project implements a probability hypothesis density tracker on data from a dual frequency identification sonar to automate the process of counting and sizing the fish crossing an insonified area. Data processing on the sonar data creates intensity images from which possible fish locations can be extracted using image processing. These locations become the input to the tracker. The probability hypothesis density tracker then solves the multiple target tracking problem and creates fish tracks from which length information is calculated using image segmentation. The algorithm is tested on data from the 2010 salmon run on the Kenai river in Alaska and compares favorably with statistical models from sub-sampling and manual measurements.

Large volume, data-driven violent conflict research is now possible using publicly available data sets. This work analyzes the predictive ability of data-derived Gaussian process models compared to a generalized linear model. Societal violence is a highly nonlinear process and the available data sets have high dimensionality that yield observation totals in the hundreds of thousands to millions. These challenges make machine learning modeling difficult without significant dimensionality reduction. We develop a computationally intensive Gaussian process modeling approach that exploits the size and complexity of the violent conflict dataset to identify appropriate basis vectors for the model. We develop our models using gridded monthly violent event counts for sub-Saharan Africa from 1980 to 2012. Our resulting Gaussian process models modestly improve the accuracy and predictive ability of existing generalized linear models. Despite this improvement, the accurate prediction of violence in sub-Saharan Africa at a relatively fine resolution spatial grid of 1 • latitude/longitude remains a challenging problem.