CS205-Artificial-Intelligence-18-Winter

Project Contents

This is the Artificial Intelligence class introduced by Dr Eamonn Keogh in UC, Riverside of 2018 Winter Quarter. This project contains:
Course material: used in class; (file: Course Material)
Homework description: and my solution; (file: Homework1)
Project 1: 8-puzzle description and my solution in C++ ; (file: Project1 8-puzzle)
Project 2: Feature Selection with Nearest Neighbor, problem description and my solution with matlab;
(file: Project2 Feature Selection)

Learning Summary

This class covers three main section :1) Search 2) Machine Learning 3)Knowledge Representation

Search:

1. Blind Search introduces the two components in search: states and operations.
   
2. Heuristc Search makes the search more quickly and accurate: A* algorithm.
   
3. Adversarial Search assume we have adversaries: Minimax Algorithm with Alpha-Beta pruning.
   

Machine Learning

1. Simple Linear Classifier: linear time to construct, constant time to use.
   
2. Nearest Neighbor Classifier: sensitve to irrelevant features.
   Feature Selection.
   
3. Decision Tree Classifier: splitting criteria: information gain, etc.
   
4. Clustering with Partition Algorithm: K-means, K-medoids, birch algorithm.
   
5. Clustering with Hierarchical Algorithm: Bottom-up: agglomerative.
   Term: dendrogram
   Distance between two clusters: single/complete/group average/words linkage.
   
6. Nearest Neighbor Clustering: best for time series.
   

Knowledge Representation

1. Propositional Logic: Truth Tables and Using Inference Rule.
   
2. First Order Logic: Substitution with GMP (not complete), resolution refutation (complete).
   Horn Form: a conjunction of atomic sentences on the left side, while a single atome on the right side.