search

FazelYU / Adaptive-Navigation

9 stars 0 forks source link

stochastic action choice #13

Closed FazelYU closed 2 years ago

FazelYU commented 3 years ago

instead of choosing the action with maximum Q-value, have a distribution over actions and choose accordingly.

FazelYU commented 2 years ago

what is the point of that? needs further research. Closed for now.