Reinforcement learning is particularly efficient when .

• A. the environment is not completely deterministic
• B. it\s often very dynamic
• C. it\s impossible to have a precise error measure
• D. all above