I can't think of a better way to relax after sitting for the Algebra PhD Qualifying Exam, than to spend an evening at Friday Jazz at The High Museum in Atlanta. Galleries and special exhibitions are open for extended hours, while live jazz is performed in the Atrium lobby every 3rd Friday of the month. The past Friday was our first time attending, but certainly not our last.
High Museum's permanent collection ranges from African to European, modern to contemporary, and folk to photographaphic art. Special exhibitions have included works of artists Leonardo DiVinci, Salvadore Dali, and others. In particular, I was attracted to a self-portrait by Chuck Close. Turns out, almost all of his work is based on a grid structure for the respresentation of an image, and in this case, himself. The result is a geometric illusion such that viewed from afar the image appears real, but as the viewer steps closer, the geometric dots and dashes and intentional tones are more apparent and the image is lost. Amazing!
Friday, July 23, 2010
Thursday, July 15, 2010
FIFA World Cup Predictions and Outcome
Congrats to the 2010 champions, ¡Viva EspaƱa! What were the chances?
Well I do not know much about soccer, but I do know that JP-Morgan, UBS and Goldman Sachs issue in depth reports every fourth year calculating the next champion based on predictive modeling. Turns out, England, the team JP-Morgan's 70 page report referred to as champs lost in the Round of 16, while UBS and Goldman reports predicted Brazil would end up on the top. C'mon... did you really need a complicated algorithm to predict that?
After reviewing the reports, the questions are: Where did Wall Street go wrong in their models (and why isn't Congress investigating it)? What was the data set, assumptions, independent and confounding variables? Is it necessary to include GDP in the calculation, and how about including team performance outside of the World Cup to the data set? Can you think of a better methodology to predict a winner? Clearly, there is no absolute truth in predictions, but we can try to improve their ability to foresee the future.
Or perhaps we should save ourselves the trouble and just depend on Paul the Octopus to call the shots?! Now I'll toot my vuvuzela to that!
Well I do not know much about soccer, but I do know that JP-Morgan, UBS and Goldman Sachs issue in depth reports every fourth year calculating the next champion based on predictive modeling. Turns out, England, the team JP-Morgan's 70 page report referred to as champs lost in the Round of 16, while UBS and Goldman reports predicted Brazil would end up on the top. C'mon... did you really need a complicated algorithm to predict that?
After reviewing the reports, the questions are: Where did Wall Street go wrong in their models (and why isn't Congress investigating it)? What was the data set, assumptions, independent and confounding variables? Is it necessary to include GDP in the calculation, and how about including team performance outside of the World Cup to the data set? Can you think of a better methodology to predict a winner? Clearly, there is no absolute truth in predictions, but we can try to improve their ability to foresee the future.
Or perhaps we should save ourselves the trouble and just depend on Paul the Octopus to call the shots?! Now I'll toot my vuvuzela to that!
Labels:
Modeling,
Prediction,
Sports
Subscribe to:
Posts (Atom)