AI That Picks Stocks Better Than the Pros
Technology Review (06/10/10) Mims, Christopher
Iona College's Robert P. Schumaker and University of Arizona's Hsinchun Chen have developed AZFinText, a system for predicting stock price fluctuations by analyzing large quantities of financial news stories along with minute-by-minute stock price data. AZFinText will buy every stock it believes will move more than one percent beyond its current price in the next 20 minutes. Schumaker and Chen minimized the amount of text the system has to parse by reducing all the financial articles into words falling into specific categories of information. The system focuses on proper nouns and combines information about their frequency with stock prices at the moment the news article is released. Using a machine-learning algorithm on historical data, AZFinText looks for correlations that can be used to predict future stock prices.
Thursday, June 10, 2010
Blog: AI That Picks Stocks Better Than the Pros
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment