Contents
1 Python for Artificial Intelligence
1.1 Why Python?
1.2 Getting Python
1.3 Running Python
1.4 Pitfalls
1.5 Features of Python
1.5.1 f-strings
1.5.2 Lists, Tuples, Sets, Dictionaries and Comprehensions
1.5.3 Functions as first-class objects
1.5.4 Generators
1.6 Useful Libraries
1.6.1 Timing Code
1.6.2 Plotting: Matplotlib
1.7 Utilities
1.7.1 Display
1.7.2 Argmax
1.7.3 Probability
1.8 Testing Code
2 Agent Architectures and Hierarchical Control
2.1 Representing Agents and Environments
2.2 Paper buying agent and environment
2.2.1 The Environment
2.2.2 The Agent
2.2.3 Plotting
2.3 Hierarchical Controller
2.3.1 Environment
2.3.2 Body
2.3.3 Middle Layer
2.3.4 Top Layer
2.3.5 Plotting
3 Searching for Solutions
3.1 Representing Search Problems
3.1.1 Explicit Representation of Search Graph
3.1.2 Paths
3.1.3 Example Search Problems
3.2 Generic Searcher and Variants
3.2.1 Searcher
3.2.2 GUI for Tracing Search
3.2.3 Frontier as a Priority Queue
3.2.4 A* Search
3.2.5 Multiple Path Pruning
3.3 Branch-and-bound Search
4 Reasoning with Constraints
4.1 Constraint Satisfaction Problems
4.1.1 Variables
4.1.2 Constraints
4.1.3 CSPs
4.1.4 Examples
4.2 A Simple Depth-first Solver
4.3 Converting CSPs to Search Problems
4.4 Consistency Algorithms
4.4.1 Direct Implementation of Domain Splitting
4.4.2 Consistency GUI
4.4.3 Domain Splitting as an interface to graph searching
4.5 Solving CSPs using Stochastic Local Search
4.5.1 Any-conflict
4.5.2 Two-Stage Choice
4.5.3 Updatable Priority Queues
4.5.4 Plotting Run-Time Distributions
4.5.5 Testing
4.6 Discrete Optimization
4.6.1 Branch-and-bound Search
5 Propositions and Inference
5.1 Representing Knowledge Bases
5.2 Bottom-up Proofs (with askables)
5.3 Top-down Proofs (with askables)
5.4 Debugging and Explanation
5.5 Assumables
5.6 Negation-as-failure
6 Deterministic Planning
6.1 Representing Actions and Planning Problems
6.1.1 Robot Delivery Domain
6.1.2 Blocks World
6.2 Forward Planning
6.2.1 Defining Heuristics for a Planner
6.3 Regression Planning
6.3.1 Defining Heuristics for a Regression Planner
6.4 Planning as a CSP
6.5 Partial-Order Planning
7 Supervised Machine Learning
7.1 Representations of Data and Predictions
7.1.1 Creating Boolean Conditions from Features
7.1.2 Evaluating Predictions
7.1.3 Creating Test and Training Sets
7.1.4 Importing Data From File
7.1.5 Augmented Features
7.2 Generic Learner Interface
7.3 Learning With No Input Features
7.3.1 Evaluation
7.4 Decision Tree Learning
7.5 Cross Validation and Parameter Tuning
7.6 Linear Regression and Classification
7.7 Boosting
7.7.1 Gradient Tree Boosting
8 Neural Networks and Deep Learning
8.1 Layers
8.1.1 Linear Layer
8.1.2 ReLU Layer
8.1.3 Sigmoid Layer
8.2 Feedforward Networks
8.3 Improved Optimization
8.3.1 Momentum
8.3.2 RMS-Prop
8.4 Dropout
8.4.1 Examples
9 Reasoning with Uncertainty
9.1 Representing Probabilistic Models
9.2 Representing Factors
9.3 Conditional Probability Distributions
9.3.1 Logistic Regression
9.3.2 Noisy-or
9.3.3 Tabular Factors and Prob
9.3.4 Decision Tree Representations of Factors
9.4 Graphical Models
9.4.1 Showing Belief Networks
9.4.2 Example Belief Networks
9.5 Inference Methods
9.5.1 Showing Posterior Distributions
9.6 Naive Search
9.7 Recursive Conditioning
9.8 Variable Elimination
9.9 Stochastic Simulation
9.9.1 Sampling from a discrete distribution
9.9.2 Sampling Methods for Belief Network Inference
9.9.3 Rejection Sampling
9.9.4 Likelihood Weighting
9.9.5 Particle Filtering
9.9.6 Examples
9.9.7 Gibbs Sampling
9.9.8 Plotting Behavior of Stochastic Simulators
9.10 Hidden Markov Models
9.10.1 Exact Filtering for HMMs
9.10.2 Localization
9.10.3 Particle Filtering for HMMs
9.10.4 Generating Examples
9.11 Dynamic Belief Networks
9.11.1 Representing Dynamic Belief Networks
9.11.2 Unrolling DBNs
9.11.3 DBN Filtering
10 Learning with Uncertainty
10.1 Bayesian Learning
10.2 K-means
10.3 EM
11 Causality
11.1 Do Questions
11.2 Counterfactual Reasoning
11.2.1 Choosing Deterministic System
11.2.2 Firing Squad Example
12 Planning with Uncertainty
12.1 Decision Networks
12.1.1 Example Decision Networks
12.1.2 Decision Functions
12.1.3 Recursive Conditioning for decision networks
12.1.4 Variable elimination for decision networks
12.2 Markov Decision Processes
12.2.1 Problem Domains
12.2.2 Value Iteration
12.2.3 Value Iteration GUI for Grid Domains
12.2.4 Asynchronous Value Iteration
13 Reinforcement Learning
13.1 Representing Agents and Environments
13.1.1 Environments
13.1.2 Agents
13.1.3 Simulating an Environment-Agent Interaction
13.1.4 Party Environment
13.1.5 Environment from a Problem Domain
13.1.6 Monster Game Environment
13.2 Q Learning
13.2.1 Exploration Strategies
13.2.2 Testing Q-learning
13.3 Q-leaning with Experience Replay
13.4 Stochastic Policy Learning Agent
13.5 Model-based Reinforcement Learner
13.6 Reinforcement Learning with Features
13.6.1 Representing Features
13.6.2 Feature-based RL learner
13.7 GUI for RL
14 Multiagent Systems
14.1 Minimax
14.1.1 Creating a two-player game
14.1.2 Minimax and - Pruning
14.2 Multiagent Learning
14.2.1 Simulating Multiagent Interaction with an Environment
14.2.2 Example Games
14.2.3 Testing Games and Environments
15 Individuals and Relations
15.1 Representing Datalog and Logic Programs
15.2 Unification
15.3 Knowledge Bases
15.4 Top-down Proof Procedure
15.5 Logic Program Example
16 Knowledge Graphs and Ontologies
16.1 Triple Store
16.2 Integrating Datalog and Triple Store
17 Relational Learning
17.1 Collaborative Filtering
17.1.1 Plotting
17.1.2 Loading Rating Sets from Files and Websites
17.1.3 Ratings of top items and users
17.2 Relational Probabilistic Models
18 Version History
Bibliography
Index