See the exercises page for assignment results!
Assignment 1: Search. The objective of the assignment is to familiarize with state-space search algorithms. Given out: January 19, Return: Thursday February 2 at 23:59
- Implement state representation for 8-puzzle: a typical representation would be a vector for 3 X 3 grid cells, indicating what tile is in each cell (the number of the tile) or if the cell is empty (denoted by 0). Include the possibility of using any number of tiles from 1 to 8.
- Implement a h-function for the 8-puzzle (the one based on Manhattan distances from the lecture.)
- Implement A*, WA*, and Greedy Best-First Search.
- You can use any programming language you prefer.
- Test each of the three algorithms with different number of tiles (1 to 8), from randomly generated initial states (have a look at the Wikipedia article on the 8-puzzle: not all initial configurations are solvable!)
- Collect statistics on the runs: number of nodes expanded and runtime.
- Report your results in a short document (text file or PDF), 1 or 2 pages.
- Return the report and all of your source code in a single archive file (tar, .gz, .Z, .bz2).
- The name of the archive file is your student ID followed by the extensions of the archive file format .tar, .gz, .Z, or .bz2.
- Email the archive file to firstname.lastname@example.org.
- The deadline is 23:59 on Thursday, February 2.
Assignment 2: Logic and Constraint Programming The objective of the assignment is to introduce to the use of logic and constraints in problem solving.
Assignment 3: Automated Planning. The objective of the assignment is to introduce model-based methods: modeling languages, model-based problem solving with general-purpose search methods.
Assignment 4: Decision-Theoretic Planning and Reinforcement Learning. The assignment introduces to decision-making with stochastic models. Deadline: March 31