recentpopularlog in

jerryking : interestingness   2

A 25-Question Twitter Quiz to Predict Retweets -
JULY 1, 2014 | NYT | Sendhil Mullainathan.

how “smart” algorithms are created from big data: Large data sets with known correct answers serve as a training bed and then new data serves as a test bed — not too differently from how we might learn what our co-workers find of the miracles of big data: Algorithms find information in unexpected places, uncovering “signal” in places we thought contained only “noise.”... the Achilles’ heel of prediction algorithms--being good at prediction often does not mean being better at creation. (1) One barrier is the oldest of statistical problems: Correlation is not causation.(2) an inherent paradox lies in predicting what is interesting. Rarity and novelty often contribute to interestingness — or at the least to drawing attention. But once an algorithm finds those things that draw attention and starts exploiting them, their value erodes. (3) Finally, and perhaps most perversely, some of the most predictive variables are circular....The new big-data tools, amazing as they are, are not magic. Like every great invention before them — whether antibiotics, electricity or even the computer itself — they have boundaries in which they excel and beyond which they can do little.
predictive_analytics  massive_data_sets  limitations  algorithms  Twitter  analytics  data  data_driven  Albert_Gore  Achilles’_heel  boundary_conditions  noise  signals  paradoxes  correlations  causality  counterintuitive  training_beds  test_beds  rarity  novelty  interestingness  hard_to_find 
july 2014 by jerryking
Prudential Research Model May Have Been a Dinosaur
June 8, 2007 | WSJ | Scott Patterson.

The decision by Prudential Financial PRU +0.70% to close its stock-research arm doesn't mean research is doomed, but it does signal an important shift. Deep-pocketed investors such as pension funds and hedge funds are hungry for exclusive, specialized research that can give them an edge over competition.

Experts say Prudential's research had become too widely distributed to draw enough interest, or dollars.

"The notion of widespread dissemination of a recommendation, that model is 40 years old," said Mike Thompson, director of research at Thomson Financial. "If you talk to the hedge funds, what they want are ideas that are actionable that not everyone gets."

The trend has given rise to independent, specialized research outfits. There are 63 independent research firms today, up from 14 in 2000, according to Thomson Financial.
equity_research  Wall_Street  investment_research  hedge_funds  pension_funds  exclusivity  nonpublic  slight_edge  proprietary  hard_to_find  novelty  interestingness  actionable_information 
february 2013 by jerryking

Copy this bookmark:

to read