Skip to content

Latest commit

 

History

1 Commit

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI — Planning Under Uncertainty with MDPs

About the course

This repository is part of my coursework for Introduction to Artificial Intelligence, a university course covering the foundations of classical AI. Topics span problem-solving as state-space search (uninformed search, A* with admissible heuristics, local search), planning under uncertainty with Markov Decision Processes, logic-based reasoning and SAT, and constraint satisfaction.

About this assignment

Where HW1 treated the environment as deterministic, HW2 lifts that assumption: actions now have probabilistic outcomes and the agent must plan over a finite horizon. The assignment builds an enumeration of the full state space, constructs the transition model (state, action) → [(next_state, probability, reward)], and implements a finite-horizon value-iteration planner. The resulting agent acts optimally by precomputing a value function over states and remaining turns, then reading off the greedy action at runtime.

Contents

  • ex2.pyBaseAgent with the state-index / transition machinery, plus the finite-horizon planner on top
  • helper.py, helper_fixed.py — state/transition utilities provided with the assignment
  • utils.py, check.py, check_fixed.py, inputs.py — supporting files and the local test harness
  • HW2.pdf — the assignment sheet

Suheil Khourieh

About

Finite-horizon planning under uncertainty with Markov Decision Processes and value iteration.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages