SE_14 AI Basics - Assessment Code Base

Description

This repository has four folders, each corresponding to one area of AI (learning, optimization, reasoning and search). Each of those folders, holds one or two projects, in which methods of the related field of AI are used to solve a problem.

The assignments for the projects is taken from the course https://cs50.harvard.edu/ai/2020/. Each of the projects had an uncompleted code base. The task was to those, to make the projects run.

Overview of projects:

Learning:

• Nim

Optimization:

• Crossword

Reasoning:

• Knights and Knaves

Search:

• Degrees
• TicTacToe

Dependencies

Install requirements.txt with:

pip3 install -r requirements.txt

Documentation of the projects

Make sure you installed the dependencies

Nim

How to run:

1. Set x in play.py (ai = train(x)) to the amount of games the AI should play befor facing you
2. Run the game with python play.py and follow the instruction in the terminal

Functions which implementation needed to be completed:

• get_q_value
• update_q_value
• best_future_reward
• choose_action

Crossword

How to run:

Run python generate.py data/structure0.txt data/words0.txt name_of_img.png

• Change the crossword layout by changing data/structure0.txt to data/structure1.txt or data/structure2.txt
• Change the set of words by changing data/words0.txt to data/words1.txt or data/words2.txt
• The last argument name_of_img.png sets the name of the saved image, leave it blank to not save the image

Functions which implementation needed to be completed:

• enforce_node_consitency
• revise
• ac3
• assignment_complete
• consistent
• order_domain_values
• select_unassinged_variable
• backtrack

Knights and Knaves

How to run: Run python puzzle.py

Functions which implementation needed to be completed:

No functions needed to be implemented for this project, only the sentences for each puzzle and the general information over the rules of the game

Degrees

How to run:

1. Run python degrees.py or python degrees_optimized.py (if you dont want to wait that long)
2. Copy and past each time a different name from large/people.csv, when the terminal asks for a name

Functions which implementation needed to be completed:

• shortest_path

Tic-tac-toe

How to run:

1. Run python runner.py
2. Follow the instructions of the window which will open

Functions which implementation needed to be completed:

• player
• actions
• result
• winner
• terminal
• utility
• minimax