def main():
# ダミー変数化のため, [0, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192]
# 入力素子数は, 16×14にされる.
ql_obj = QLearning(16, 4, dummy=False)
max_score = 0
score_list = []
for i in range(10000):
score, result = play(ql_obj)
# Q-learning
ql_obj.train(result)
score_list.append(score)
# print weight
data = [[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 2], [0, 0, 0, 2]]
output_vec = ql_obj.get_q_values(data)
print i, numpy.mean(score_list), max(score_list), output_vec
def main():
# ダミー変数化のため, [0, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192]
# 入力素子数は, 16×14にされる.
ql_obj = QLearning(16, 4, dummy=False)
max_score = 0
score_list = []
for i in range(10000):
score, result = play(ql_obj)
# Q-learning
ql_obj.train(result)
score_list.append(score)
# print weight
data =[[0,0,0,0], [0,0,0,0], [0,0,0,2], [0,0,0,2]]
output_vec= ql_obj.get_q_values(data)
print i, numpy.mean(score_list) , max(score_list), output_vec
