Winners of Neukom Scholarships for the 2008-2009 academic year were:
The Green Lite Project aims to show people the impact that their decisions make on the environment through displays and animations. Originally, Green Lite monitors measuring energy use had been installed at several locations on campus. This new project expanded those efforts to give residents feedback for water usage in addition to power. Zhao created an animation that provided that feedback. Video
Recent events in worldwide credit markets have demonstrated the striking necessity for understanding and predicting out-of-equilibrium dynamics within such systems. Yet due to the enormous complexity of these large networks of heterogeneous agents, many of the standard analytical tools designed to analyze financial markets are sorely lacking. The development of new computational models of credit networks, therefore, may prove extraordinarily helpful. To that end, this project sought to design and implement a novel agent-based simulation of credit transactions within a network structure.
Spring term '09 was the full launch of Green Lite on campus, alongside the Dartmouth Energy Campaign. Schnippering continued her work as an important part of the project team by finalizing details in the desktop widget; she was able to test it in its real setting distributed across campus. The research component that she continued to address was studying and resolving how to form the displays to have the intended effect on students, which happens largely through computational methods.
Computation of the electrical capacitance of a geometrically-complex conducting body is a common engineering problem, occurring for instance in designing micro-chip interconnects or evaluating capacitances of biomembranes. The self-capacitance of a unit cube is a paradigm problem: as no analytical expression is known, a numerical approximation is crucial. Our project was to implement a new spectrally accurate method, for which the accuracy converges exponentially as (much faster than the weak algebraic convergence observed by the Monte Carlo method), where N is the number of degrees of freedom in the problem.
