User:Miami Thrice/interests

Data Driven Modeling & Analysis
Data Driven Modeling and Analysis is the intelligent extraction of information from a noisy, transformed, and often overwhelmingly large dataset. These methods are developed with the use of prior knowledge, weak assumptions, and the incorporation of a process’s underlying physics.

Data Driven Modeling & Analysis Team at LANL

Robust Uncertainty Principles: Exact Signal Reconstruction from Highly Incomplete Frequency Information by Terence Tao, Emmanuel Candes, and Justin Romberg.

Economics Engineering
Avoiding the pitfalls when economics shifts from science to engineering by Hal Varian

The Economist as Engineer: Game Theory, Experimentation, and Computation as Tools for Design Economics by Alvin Roth

Pairwise kidney exchange by Alvin E. Roth, Tayfun Sönmezc, M. Utku Ünver

Market Intelligence
Detecting Cartels by Joseph E. Harrington, Jr.

Technical Solutions for Democracy
Social choice theory is a subdiscipline within economics that examines the mechanics of democracy. While the founding fathers of American democracy were

Electronic Voting
Past elections have show serious weaknesses and abuses of electronic voting machines - those by Dieblod in particular. Novel studies in cryptography have formalized and addressed the weaknesses in our voting system. Voters can have mathematical assurance of correctness for large-scale elections, without compromising privacy. When we adopt a cryptographic methodology to electronic voting, we can resolve even the most closely contested of elections without the blemished the results caused by allegations of ballot-box manipulation. Good introductions include:

"A practical secret voting scheme for large scale elections", by Atsushi Fujioka, Tatsuaki Okamoto, and Kazuo Ohta. Advances in Cryptology -- AUSCRYPT '92, pp 244-251, 1992. is the universally cited paper that formalizes the framework of the proper.

Bruce Schneier's Applied Cryptography has an excellent chapter on the subject

"Efficient Receipt Free Voting Based on Homomorphic Encryption" by Martin Hirt and Kazue Sako is a more recent example of what a solution might resemble. 

We're so accustomed to seeing the shoddy work of our domestic voting technology companies (who go out of their way to disable the audit logs on their Microsoft Access backend) that we internalize this belief that a truly reliable electronic voting system is unattainable and the only protection is a paper trail. With the power of math, we can do much better than that.

Algorithmic Allocation of Congressional Districts
The current allocation of our congressional districts is mostly done by political parties. This results in an allocation of districts that have been manipulated to serve the interest of the majority party. The notorious 2003 Texas redistricting was a prime example of this abuse. The Republican party of Pennsylvania even contracted researchers at my alma mater to develop optimally gerry-mandered districts. A fair allocation of congressional districts could be performed by a simple algorithm incorporating only the geographical population distribution. There are many candidate algorithms but RangeVoting has developed one example solution: an iterative method that divides the state by selecting the shortest dividing line that will equitably split the population..

Alternative Voting Systems
The plurality voting system is the only sensible democratic system when there are exactly two candidates. The plurality voting system, however, heavily disadvantages third party candidacies. Unlike our difficulties with electronic voting, no perfect alternative voting system can exist. (see Arrow's Theorem). However, some systems are better than others and many are better than the one we currently have. In my amateur opinion, condorcet methods may be the way to go.

National Popular Vote
Most Americans would prefer that presidential elections were settled the same way we elect every other elected official in the United States: a simple popular vote. We arrived at the current system not by the deliberate intentions of our founding fathers but by historical accident. Most believe we cannot switch to the national popular vote without an unlikely amendment to the constitution. However, a Stanford Computer Science professor, John R. Koza, developed a novel plan for implementing a national popular vote through the use of an interstate compact. The National Popular Vote Interstate Compact already has one sixth of the electoral college votes needed to becoming law of the land and revolutionizing our elections. You can support them at nationalpopularvote.com or read Count 'Em by Hendrik Hertzberg if you need more convincing.