Introduction - If you have any usage issues, please Google them yourself
In this book, we explored a computational method of learning from interaction. We do not directly analyze how people or animals learn, but explore idealized learning situations and evaluate the efficiency of various learning methods. That is to say, we adopt the angle of artificial intelligence researchers or engineers. We are exploring ways to design machines that are exceptionally efficient in these areas and that can solve problems in science or economics. These designs are evaluated through data analysis and computational experiments. We call this method reinforcement learning, which focuses more on goal-oriented interactive learning than other methods.