This project uses the cvxpy library to solve the optimization problem.
You can read the official documentation here.
Dhruvee Birla and myself.
This assignment was done as a part of the Machine, Data and Learning course, Spring 2021.
The goal of this exercise is to understand Markov Decision Processes using Mathematical Programming.
Here, we work on the same MDP problem as used in the value iteration exercise, with one change: When MM’s health reaches 0, IJ quest finished but gets zero reward. All other costs and rewards are the same.
In this exercise, we formulate the problem a LP and solve it. The main logic is there in the notebook, and the script generates the output. The approach, observations and conclusions have been summarised in the report.