O5-B Deep Learning for Improving Stock Market Investment

PI: Zhibin (Ben) Yang

The objective of this project is to create data-driven deep learning models that

  • Optimizes the institutional investor’s investment strategy in the stock market
  • Provides decision support for stock pick
  • Are “accountable”
  • Are flexible enough to properly accommodate heterogeneous data

We will create a Deep Learning Network with

  • Hierarchical structure that reflects and preserves the multi-tier thinking of the investment strategy
  • Modularized design that
  • Accommodates input data with heterogeneous frequencies
  • Preserves conceptual distinction between “factors”