These are my solutions to edX Edge Artificial Intelligence - Berkeley CS188-SU16 (Summer 2016) course instructed by Davis Foote and Jacob Andreas. I just want to thank them for this amazing course and for those challenging projects 
.
- Project 0: Tutorial
- Project 1: Search
- Project 2: Multiagent
- Project 3: Reinforcement Learning
- Project 4: Bayes Nets (Not available on CS188 website)
- Project 5: Tracking
- Project 6: Classification
For projects 0 to 5, it is straightforward to run them locally since they only use Python Standard Library. Just make sure you are using Python 2. However, for the last project, tensorflow library is needed and it might be tricky to install it. For that reason, a Docker image which has all the required dependencies already installed is provided. Check how to use it in Docker section.
If you want to run a single question from a project, use the following commands. Note that QUESTION is q1, q2, up to the number of questions of the project.
git clone https://github.com/EikaNN/Berkeley-AI-CS188.git
cd Berkeley-AI-CS188
cd project1-search
python autograder.py [-q QUESTION]If you want to run multiple projects, or all the questions from one project, you can use the main.py script that I have implemented. Note that PROJECT is either 0, 1, 2, 3, 4 or 5.
git clone https://github.com/EikaNN/Berkeley-AI-CS188.git
cd Berkeley-AI-CS188
python main.py [-p PROJECT]To run the projects with Docker (again, PROJECT is either 0, 1, 2, 3, 4 or 5):
git clone https://github.com/EikaNN/Berkeley-AI-CS188.git
cd Berkeley-AI-CS188
./docker.sh [-p PROJECT]In this section the algorithms that were implemented in each project are listed. For each project, there is a commit with the blank project followed by a commit with all the changes that were made. This way, it is easy to identify what has been implemented. All projects are completed with the maximum score.
Introduction to the grading environment.
Provisional grades
==================
Question q1: 1/1
Question q2: 1/1
Question q3: 1/1
------------------
Total: 3/3
Implementation of DFS (Depth First Search), BFS (Breadth First Search), UCS (Uniform Cost Search) and A* search with heuristics.
Provisional grades
==================
Question q1: 3/3
Question q2: 3/3
Question q3: 3/3
Question q4: 3/3
Question q5: 3/3
Question q6: 3/3
Question q7: 5/4
Question q8: 3/3
------------------
Total: 26/25
Implementation of Minimax, Alpha-Beta Pruning and Expectimax.
Provisional grades
==================
Question q1: 4/4
Question q2: 5/5
Question q3: 5/5
Question q4: 5/5
Question q5: 6/6
------------------
Total: 25/25
Implementation of Q-learning.
Provisional grades
==================
Question q1: 4/4
Question q2: 1/1
Question q3: 5/5
Question q4: 1/1
Question q5: 3/3
Question q6: 4/4
Question q7: 2/2
Question q8: 1/1
Question q9: 1/1
Question q10: 3/3
------------------
Total: 25/25
Implementation of variable elimination and value-of-perfect-information computations for Bayes Nets.
Provisional grades
==================
Question q1: 3/3
Question q2: 3/3
Question q3: 5/5
Question q4: 4/4
Question q5: 4/4
Question q6: 4/4
Question q7: 1/1
Question q8: 4/4
------------------
Total: 28/28
Implementation of exact and approximante inference using Bayes Nets.
Provisional grades
==================
Question q1: 2/2
Question q2: 3/3
Question q3: 3/3
Question q4: 2/2
Question q5: 2/2
Question q6: 3/3
Question q7: 3/3
Question q8: 1/1
Question q9: 3/3
Question q10: 3/3
------------------
Total: 25/25
Implementation of the perceptron algorithm and gradient descent to train neural network classifiers.
Provisional grades
==================
Question q1: 4/4
Question q2: 1/1
Question q3: 1/1
Question q4: 2/2
Question q5: 1/1
Question q6a: 2/2
Question q6b: 1/1
Question q6c: 1/1
Question q7: 1/1
Question q8: 2/2
Question q9: 5/5
Question q10: 4/4
------------------
Total: 25/25