Use the exercises for your enjoyment and learning and to gain additional experience. Deep learning and deep reinforcement learning are very much areas where your knowledge will only improve by working with the examples. Don't expect to be a natural with training agents; it takes a lot of trial and error. Fortunately, the amount of experience we need is not as much as our poor agents require but still expect to put some time in.
- Open example Chapter_8_REINFORCE.py back up and alter the hyperparameters to see what effect this has on training.
- Open example Chapter_8_ActorCritic.py back up and alter the hyperparameters to see what effect this has on training.
- Open example Chapter_8_DDPG.py back up and alter the hyperparameters to see what effect this has on training.
- How can you convert the REINFORCE or ActorCritic examples to use continuous action spaces? Attempt...
