Development of an agent-based simulation model in combination with reinforcement learning in Python using Mesa library - At the beginning of an episode, 10 plants (as agents) are planted - Plants must grow for 10 days (steps) before they can be harvested. - Each plant has a 10% chance of dying every day. - A new (fresh) plant can be bought every day (cost $10) to be planted - The aim is to harvest 10 plants that each grew for 10 days. When the goal is reached, there is a reward of $20 per plant harvested and the episode ends - Each day of the episode costs $5 - Reinforcement learning is now used to find a strategy when to plant new trees that minimizes total costs
