Deliverables

• Open Source implementation of an autonomous learning system that can be used in complex, dynamic, environments — See Road Map for a list of planned features
• Publications on autonomous learning versus learning from examples
• Publications on the performance of relatively simple probabilistic models on complex environments
• Publications on the theoretical limits of a universal autonomous learning system
• Publications on the (theoretical) performance of autonomous learning systems in non-zero-sum games
• Publications on the applicability of autonomous learning systems in operations research
• Publications on efficient algorithms for
  • calculating the lengths of two-part codes
  • finding candidate models
  • finding and evaluating candidate policies

Visit the LEIA project page on GitHub for project details.