modelling artificial creature behavior and emotions with CLA
- python2
- NuPIC
- utility encoder branch to NuPIC
- (mayavi - via pip)
- action learning - learning prerequisities, effects and "meaning" of action. Imagine a baby learning to "see" or control its legs.
- emotions - emotions to represent inner state, goals, uncontious reactions (fear, hunger); using
utility-encoder
- behavior - How it can simplify programming when sub-goals shift priorities automatically according to inner/outer/random conditions.
-
simple utility map of a terrain, eucl. distance to target:
python alife/experiments/utility_map/utility_map.py
-
behavior: goal is to reach target, but with each step, the agent becomes more and more hungry, it also perceives food located on a map and plans accordingly.
python alife/experiments/behavior/utility_map.py
and target and hunger combined...
python alife/experiments/behavior/random_walk_map.py