Optimizing Decision-Making via MDP Analysis
Budget: £250 – £750 GBP
I'm looking to solve a Markov Decision Problem (MDP) with the aim to optimize a decision-making process. The project will primarily involve a theoretical study of optimal choice. At the current moment, there is not data involved (potential for a follow-up project based on satisfaction). A detailed pdf will be shared after initial discussion.
Key Tasks:
- Review existing code.
- Utilize Bellman equations and value function iteration to analyse the MDP.
- Important: the state space dynamics are partially stochastic and partially deterministic (key challenge).
- Explore policy iteration and Monte Carlo methods as alternative strategies.
- Deliver the final code.
Ideal Skills and Experience:
- Strong background in MDPs and reinforcement learning.
- Proficiency with Bellman equations, value function iteration, policy iteration, and Monte Carlo methods.
- Experience in conducting theoretical studies and making practical recommendations based on findings.
Key Tasks:
- Review existing code.
- Utilize Bellman equations and value function iteration to analyse the MDP.
- Important: the state space dynamics are partially stochastic and partially deterministic (key challenge).
- Explore policy iteration and Monte Carlo methods as alternative strategies.
- Deliver the final code.
Ideal Skills and Experience:
- Strong background in MDPs and reinforcement learning.
- Proficiency with Bellman equations, value function iteration, policy iteration, and Monte Carlo methods.
- Experience in conducting theoretical studies and making practical recommendations based on findings.