Next:
DS Advantage 1: No
Up:
Sequential Decision Making Based
Previous:
Introduction
Advantages of Direct Search
Subsections
DS Advantage 1: No States
DS Advantage 2: No Markovian Restrictions
DS Advantage 3: Straight-forward Hierarchical Credit Assignment
DS Advantage 4: Non-Hierarchical Abstract Credit Assignment
DS Advantage 5: Metalearning Potential
Advantage 6: Exploring Limited Spatio-Temporal Predictability
Summary
Juergen Schmidhuber
2003-02-19
Back to
Reinforcement Learning and POMDP page
