Hi,
Time 24:48, why use Markov Chain instead of Monte Carlo?
Course Name: Deep Reinforcement Learning in Trading, Section No: 7, Unit No: 6, Unit type: WaterMarkVideo
Hello,
Markov Chains let models the step-by-step transitions between states. They focus on how the system evolves over time based on the current state and action. This makes them ideal for reinforcement learning, where the agent needs to learn and improve as it goes, not just at the end.
Monte Carlo, on the other hand, waits until the end of a full episode before it can update anything. That’s fine for some tasks, but in many real-world problems like trading, you don’t always have clear episode endings. You need something that can learn from each step.
That’s why the Markov approach was used, it’s more flexible and fits better with how RL works in dynamic environments.