Monday, March 4, 2013

The Signal and the Noise: Why So Many Predictions Fail — but Some Don't


The Signal and the Noise: Why So Many Predictions Fail — but Some Don't by Nate Silver (Author). Nate Silver constructed an modern system for predicting baseball performance, predicted the 2008 election within a hair’s breadth, and became a nationwide sensation as a blogger-all by the time he was thirty. The New York Occasions now publishes FiveThirtyEight.com, where Silver is among the nation’s most influential political forecasters.

 Drawing on his personal groundbreaking work, Silver examines the world of prediction, investigating how we can distinguish a real signal from a universe of noisy data. Most predictions fail, usually at great value to society, as a result of most of us have a poor understanding of probability and uncertainty. Each specialists and laypeople mistake more confident predictions for extra accurate ones. But overconfidence is often the explanation for failure. If our appreciation of uncertainty improves, our predictions can get better too. That is the “prediction paradox”: The extra humility we've got about our means to make predictions, the extra profitable we might be in planning for the future by Nate Silver (Author).


This book explains the unerring accuracy for Nate SIlver's election predictions using Bayesian statistics. The BEST a part of the guide for me was that I lastly understand Bayes' analysis. I used quite a number of sophisticated statistical tools in my work (retired as reliability physics professional for semiconductor devices, aka chips), however I was by no means capable of grasp Bayes Theorem until now. Wikipedia's "tutorial" was far too sophisticated even for a PhD, however Nate offered a easy model that a layman can perceive ... and he did it using a hilarious instance (look for "dishonest"). In reality, I am so impressed with Bayes' evaluation that I'm eager about writing a corollary to my two best technical papers grafting a Bayesian view. Returning to the election prediction problem, consider that every ballot of 1000 folks has a sampling error of +-5%, simply derived from Poisson statistics. Nevertheless, when one pools the outcomes from say 25 polls (and removes bias), the pattern size is elevated by 25-fold, which reduces the sampling error by 5-fold, all the way down to +-1%. Thus, one can make assured predictions over variations FAR smaller than the same old sampling error. When one combines Bayesian pooling with a state-by-state analysis, one can make astonishingly accurate predictions ... Nate predicted ALL 50 states appropriately, so his electoral count was exactly on reality as effectively when fractional electoral counts are eliminated. 
Purchase the book as it is academic and fun to read.

A nicely written ebook with case research throughout many areas of prediction. While not shying away from a few of the math, the e book continues to be accessible and entertaining to common readers in addition to we that make our living making an attempt to make strong predictions about future events. He consists of many hints and greatest practices which can assist to raise our collective predictive capabilities.

I highly advocate buying this book. It's going to certainly provide you with insights on why his strategies produced such an accurate predictions of the 2012 elections. It isn't magic, relatively a road map to the future. 

 The Signal and the Noise: Why So Many Predictions Fail — but Some Don't 
 Nate Silver (Author)
 544 pages
 Penguin Press HC, The; 1 edition (September 27, 2012)


No comments:

Post a Comment