We attempted to solve the optimal route problem in the city of Mumbai, India using various algorithms in varying levels of difficulty.
- Built a linear programming-based pathfinder across Mumbai using 4 transits, improving ETA accuracy by 20%
- Developed a stochastic version by modeling the city as an MDP with 1M+ route permutations, implementing a user-defined time and cost-based reward function to compute optimal route policy using policy, value iteration
A detailed poster presentation can be found here - https://drive.google.com/file/d/1EVFKHjt9yr1wmeOcsbGK2ySje0c7orKv/view