Agile Artificial Intelligence in Pharo: Implementing Neural Networks, Genetic Algorithms, and Neuroevolution

Table of Contents
About the Author
About the Technical Reviewer
Acknowledgments
Introduction
Part I: Neural Networks
	Chapter 1: The Perceptron Model
		1.1 Perceptron as a Kind of Neuron
		1.2 Implementing the Perceptron
		1.3 Testing the Code
		1.4 Formulating Logical Expressions
		1.5 Handling Errors
		1.6 Combining Perceptrons
		1.7 Training a Perceptron
		1.8 Drawing Graphs
		1.9 Predicting and 2D Points
		1.10 Measuring the Precision
		1.11 Historical Perspective
		1.12 Exercises
		1.13 What Have We Seen in This Chapter?
		1.14 Further Reading About Pharo
	Chapter 2: The Artificial Neuron
		2.1 Limit of the Perceptron
		2.2 Activation Function
		2.3 The Sigmoid Neuron
		2.4 Implementing the Activation Functions
		2.5 Extending the Neuron with the Activation Functions
		2.6 Adapting the Existing Tests
		2.7 Testing the Sigmoid Neuron
		2.8 Slower to Learn
		2.9 What Have We Seen in This Chapter?
	Chapter 3: Neural Networks
		3.1 General Architecture
		3.2 Neural Layer
		3.3 Modeling a Neural Network
		3.4 Backpropagation
			3.4.1 Step 1: Forward Feeding
			3.4.2 Step 2: Error Backward Propagation
			3.4.3 Step 3: Updating Neuron Parameters
		3.5 What Have We Seen in This Chapter?
	Chapter 4: Theory on Learning
		4.1 Loss Function
		4.2 Gradient Descent
		4.3 Parameter Update
		4.4 Gradient Descent in Our Implementation
		4.5 Stochastic Gradient Descent
		4.6 The Derivative of the Sigmoid Function
		4.7 What Have We Seen in This Chapter?
		4.8 Further Reading
	Chapter 5: Data Classification
		5.1 Training a Network
		5.2 Neural Network as a Hashmap
		5.3 Visualizing the Error and the Topology
		5.4 Contradictory Data
		5.5 Classifying Data and One-Hot Encoding
		5.6 The Iris Dataset
		5.7 Training a Network with the Iris Dataset
		5.8 The Effect of the Learning Curve
		5.9 Testing and Validation
		5.10 Normalization
		5.11 Integrating Normalization into the NNetwork Class
		5.12 What Have We Seen in This Chapter?
	Chapter 6: A Matrix Library
		6.1 Matrix Operations in C
		6.2 The Matrix Class
		6.3 Creating the Unit Test
		6.4 Accessing and Modifying the Content of a Matrix
		6.5 Summing Matrices
		6.6 Printing a Matrix
		6.7 Expressing Vectors
		6.8 Factors
		6.9 Dividing a Matrix by a Factor
		6.10 Matrix Product
		6.11 Matrix Subtraction
		6.12 Filling the Matrix with Random Numbers
		6.13 Summing the Matrix Values
		6.14 Transposing a Matrix
		6.15 Example
		6.16 What Have We Seen in This Chapter?
	Chapter 7: Matrix-Based Neural Networks
		7.1 Defining a Matrix-Based Layer
		7.2 Defining a Matrix-Based Neural Network
		7.3 Visualizing the Results
		7.4 Iris Flower Dataset
		7.5 What Have We Seen in This Chapter?
Part II: Genetic Algorithms
	Chapter 8: Genetic Algorithms
		8.1 Algorithms Inspired from Natural Evolution
		8.2 Example of a Genetic Algorithm
		8.3 Relevant Vocabulary
		8.4 Modeling Individuals
		8.5 Crossover Genetic Operations
		8.6 Mutation Genetic Operations
		8.7 Parent Selection
		8.8 Evolution Monitoring
		8.9 The Genetic Algorithm Engine
		8.10 Terminating the Algorithm
		8.11 Testing the Algorithm
		8.12 Visualizing Population Evolution
		8.13 What Have We Seen in This Chapter?
	Chapter 9: Genetic Algorithms in Action
		9.1 Fundamental Theorem of Arithmetic
		9.2 The Knapsack Problem
			9.2.1 The Unbounded Knapsack Problem Variant
			9.2.2 The 0-1 Knapsack Problem Variant
			9.2.3 Coding and Encoding
		9.3 Meeting Room Scheduling Problem
		9.4 Mini Sodoku
		9.5 What Have We Seen in This Chapter?
	Chapter 10: The Traveling Salesman Problem
		10.1 Illustration of the Problem
		10.2 Relevance of the Traveling Salesman Problem
		10.3 Naive Approach
		10.4 Adequate Genetic Operations
		10.5 The Swap Mutation Operation
		10.6 The Ordered Crossover Operation
		10.7 Revisiting Our Large Example
		10.8 What Have We Seen in This Chapter?
	Chapter 11: Exiting a Maze
		11.1 Encoding the Robot’s Behavior
		11.2 Robot Definition
		11.3 Map Definition
		11.4 Example
		11.5 What Have We Seen in This Chapter?
	Chapter 12: Building Zoomorphic Creatures
		12.1 Modeling Join Points
		12.2 Modeling Platforms
		12.3 Defining Muscles
		12.4 Generating Muscles
		12.5 Defining the Creature
		12.6 Creating Creatures
			12.6.1 Serialization and Materialization of a Creature
			12.6.2 Accessors and Utility Methods
		12.7 Defining the World
		12.8 Cold Run
		12.9 What Have We Seen in This Chapter?
	Chapter 13: Evolving Zoomorphic Creatures
		13.1 Interrupting a Process
		13.2 Monitoring the Execution Time
		13.3 The Competing Conventions Problem
		13.4 The Constrained Crossover Operation
		13.5 Moving Forward
		13.6 Serializing the Muscle Attributes
		13.7 Passing Obstacles
		13.8 Climbing Stairs
		13.9 What Have We Seen in This Chapter?
Part III: Neuroevolution
	Chapter 14: Neuroevolution
		14.1 Supervised, Unsupervised Learning, and  Reinforcement Learning
		14.2 Neuroevolution
		14.3 Two Neuroevolution Techniques
		14.4 The NeuroGenetic Approach
		14.5 Extending the Neural Network
		14.6 NeuroGenetic by Example
		14.7 The Iris Dataset
		14.8 Further Reading About NeuroGenetic
		14.9 What Have We Seen in This Chapter?
	Chapter 15: Neuroevolution with NEAT
		15.1 Vocabulary
		15.2 The Node Class
		15.3 Different Kinds of Nodes
		15.4 Connections
		15.5 The Individual Class
		15.6 Species
		15.7 Speciation
		15.8 The Crossover Operation
		15.9 Abstract Definition of Mutation
		15.10 Structural Mutation Operations
			15.10.1 Adding a Connection
			15.10.2 Adding a Node
		15.11 Non-Structural Mutation Operation
		15.12 Logging
		15.13 NEAT
		15.14 Visualization
		15.15 The XOR Example
		15.16 The Iris Example
		15.17 What Have We Seen in This Chapter?
	Chapter 16: The MiniMario Video Game
		16.1 Character Definition
		16.2 Modeling Mario
		16.3 Modeling an Artificial Mario Player
		16.4 Modeling Monsters
		16.5 Modeling the MiniMario World
		16.6 Building the Game’s Visuals
		16.7 Running MiniMario
		16.8 NEAT and MiniMario
		16.9 What Have We Seen in This Chapter?
Afterword
Last Words
Index