Exploration Entropy for Reinforcement Learning

Figure 7

State Entropy of all states at the (a) 10th, (b) 20th, (c) 40th, (d) 60th, (e) 100th, (f) 200th, (g) 300th, (h) 800th, and (i) 1000th iteration with Softmax strategy in Maze B.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)