The problem with investing based on pattern recognition

A famous story in artificial intelligence is how the US military developed algorithms to determine whether an image had a tank in it. They used a standard machine learning method: feed the computer a training set of photos, some of which had tanks in them and some of which didnt, and let algorithms identify which features in the photos correlated to tanks being shown. Since the features the computer identified were embedded in complicated mathematical equations, no one could figure out what it was really doing and therefore why it stopped working.Eventually someone realized that in the training set, all of the images with tanks were taken on a cloudy day, and all the images without tanks were taken on a sunny day. In the context of markets, it can cause companies and sectors with the right patterns to be overvalued, and ones with the wrong patterns to be undervalued.

Author: Publication Date: Crawl Date: Source: Topic: Contributor: Type: Language: Format:
March 07, 2012 March 11, 2012 Business Insider via Google News Applications MachineLearning NewsFinder Text English HTML

Whitelist words:

  1. pattern recognit: 3 occurrences
  2. artifici intellig: 1 occurrence
  3. machin learn: 1 occurrence

Categorization (need >= 0.5 to match):

  1. AIOverview: 0.169573
  2. Agents: 0.00881608
  3. Applications: 0.80282
  4. CognitiveScience: 0.0262083
  5. Education: 0.00274227
  6. Ethics: 0.00275335
  7. Games: 0.0325192
  8. History: 0.0556831
  9. Interfaces: 0.00963408
  10. MachineLearning: 0.954268
  11. NaturalLanguage: 0.00484034
  12. Philosophy: 0.000940789
  13. Reasoning: 0.15943
  14. Representation: 0.0472249
  15. Robots: 0.0281981
  16. ScienceFiction: 0.0928278
  17. Speech: 0.0114124
  18. Systems: 0.0202165
  19. Vision: 0.178406

Duplicates (threshold=0.17):


  1. Published
AAAI   Recent Changes   Edit   History   Print   Contact Us
Page last modified on March 12, 2012, at 12:00 AM