I would like to create an agent-based simulation model in which a single agent, a person generated at the space's boundary, aims to reach a predefined destination while avoiding nearby agents and obstacles, which will be walls. Which machine learning algorithm will be most suitable for this model and what dataset will be required?

  • $\begingroup$ There are some robotics algorithms for path finding while avoiding obstacles. If you can formulate your environment as a state space, you may also be able to use a state-space search algorithm. Alternatively, you may also be able to use RL. It could be useful to give more details about your space (for example, an image). Is the goal just to avoid obstacles or do you also want to get to the goal as quickly as possible, etc $\endgroup$
    – nbro
    Commented Jul 4 at 10:47


You must log in to answer this question.