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

  

  Utility as a distance to target.

• 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

  

  Locations of food - attractors, influence changes with hunger level.

  and target and hunger combined...

  

  Overall utility of target & food (*.

python alife/experiments/behavior/random_walk_map.py

A single random walk with utility, food, target,...