Reinforcement Learning in Stock Market Simulation
Budget: $30 – $250 USD
it is a collaborative artificial intelligence project:
Goal: The goal of your project is to implement your own stock market simulation environment, adhering to the OpenAI Gym specification, and then develop an RL agent to solve it. While you will be provided with general guidance and a code skeleton, you are encouraged to write your own code and explore alternative approaches. This project is not a typical homework assignment; it is an open-ended project. You are expected to conduct your own research, address challenges independently, and develop a deeper understanding of reinforcement learning and environment design.
Task Breakdown: The project is divided into two main parts:
Part One: Implement the basic environment with a simple action space of 3 actions.
Part Two: Expand the environment to include a pool of money and additional complexities.
It is recommended to start with Part One, but you may proceed with Part Two directly and treat Part One as a subcase of Part Two. Regardless of your approach, the environment must adhere to Gym specifications and be capable of running the provided models.
Code Flexibility: You have the flexibility to modify the provided code or write your own entirely. However, your final implementation must comply with the Gym specification and be able to run with the provided models. Aim to achieve a positive reward on your initial implementation (without the pool of money) to demonstrate basic functionality.
Goal: The goal of your project is to implement your own stock market simulation environment, adhering to the OpenAI Gym specification, and then develop an RL agent to solve it. While you will be provided with general guidance and a code skeleton, you are encouraged to write your own code and explore alternative approaches. This project is not a typical homework assignment; it is an open-ended project. You are expected to conduct your own research, address challenges independently, and develop a deeper understanding of reinforcement learning and environment design.
Task Breakdown: The project is divided into two main parts:
Part One: Implement the basic environment with a simple action space of 3 actions.
Part Two: Expand the environment to include a pool of money and additional complexities.
It is recommended to start with Part One, but you may proceed with Part Two directly and treat Part One as a subcase of Part Two. Regardless of your approach, the environment must adhere to Gym specifications and be capable of running the provided models.
Code Flexibility: You have the flexibility to modify the provided code or write your own entirely. However, your final implementation must comply with the Gym specification and be able to run with the provided models. Aim to achieve a positive reward on your initial implementation (without the pool of money) to demonstrate basic functionality.