This is an interesting piece from The Economist - worth a read if you can access it (may be gated):

Hedge funds embrace machine learning-up to a point

  • left to their own devices, machine-learning techniques are prone to "overfit"
  • This is especially true of financial data ... because of their comparative paucity. Share-price time series going back decades still contain far less information than, say, the image data used to used to train Facebook's facial-recognition algorithms

Its only a quickie article - perfect for a Monday!