کتاب هوش مصنوعی: اصول و عمل (۲۰۲۵)

Preface
۱.۱ Using This Book
۱.۲ Programming
۱.۳ About the Author
Contents
I: Introducing Artificial Intelligence
۱: The Pre-History of Artificial Intelligence
۱.۱ Mary Shelley, Frankenstein, and Prometheus
۱.۲ The Age of Rationalism
۱.۳ The Empiricist Tradition
۱.۴ Immanuel Kant: Bridging the Rationalist/Empiricist Viewpoints
۱.۵ The Reverend Thomas Bayes
۱.۶ The Mathematical Foundations for Artificial Intelligence
۱.۷ American Pragmatism
۱.۸ The Turing Test: Can a Machine Be “Intelligent”?
۱.۹ The 1956 Dartmouth Summer Workshop
۱.۱۰ Summary
۱.۱۱ Exercises
۲: Computing, Representations, and Definitions of Artificial Intelligence
۲.۱ Artificial Intelligence: Attempting a Definition
۲.۲ Computer-Based Representations of the World
۲.۳ The General Themes of Current AI Practice
۲.۳.۱ Symbol-Based AI
۲.۳.۲ Neural Networks
۲.۳.۳ Genetic and Emergent AI
۲.۳.۴ Probabilistic Models
۲.۴ Summary of Part I and an Introduction to Part II
۲.۵ Exercises
II: Symbol-Based AI: Foundations
۳: The State Space, Finite State Machines, and Artificial Life
۳.۱ Graph Theory: The Origins of the State Space Representation
۳.۲ The State Space Representation
۳.۳ The Finite State Machine
۳.۴ Artificial Life: The Emergence of Complexity
۳.۴.۱ Artificial Life: Examples
۳.۴.۲ Extensions of A-Life
۳.۵ Summary
۳.۶ Exercises
۴: Searching the State Space
۴.۱ Data-Driven and Goal-Driven Search
۴.۲ Implementing Graph Search with the Backtrack Algorithm
۴.۳ Breadth-First and Depth-First Search
۴.۳.۱ The Breadth-First Algorithm
۴.۳.۲ The Depth-First Search Algorithm
۴.۳.۳ Depth-First Search with Iterative Deepening
۴.۴ Extending Search Strategies to and/or Graphs
۴.۵ Summary
۴.۶ Exercises
۵: Heuristic Search
۵.۱ An Introduction to Heuristics
۵.۲ Hill-Climbing
۵.۳ Best-First Search
۵.۳.۱ The Best_First_Search Algorithm
۵.۳.۲ Adding a Depth Measure to Best-First Search
۵.۴ Genetic Algorithms: Evaluating Multiple “Best” States
۵.۵ In Summary
۵.۶ Exercises
۶: Heuristics: 2-Person Games and Theoretical Constraints
۶.۱ The Minimax Procedure on Exhaustively Searchable Graphs
۶.۲ Using Minimax to a Fixed Ply Depth
۶.۳ The Alpha-Beta Procedure
۶.۴ Multi-Person Games, Admissibility, Monotonicity, and Informedness
۶.۴.۱ Using Heuristics, the Zero-Sum Hypothesis, and Multi-Person Games
۶.۴.۲ Admissibility Measures
۶.۴.۳ Monotonicity
۶.۴.۴ When One Heuristic Is “Better” Than Another: Informedness
۶.۵ Heuristics and Complexity
۶.۶ Summary
۶.۷ Exercises
III: The Propositional and Predicate Calculi and Resolution-Based Reasoning
۷: Introduction to the Propositional and Predicate Calculi
۷.۱ The Propositional Calculus
۷.۱.۱ Symbols and Sentences
۷.۱.۲ The Semantics of Propositional Calculus
۷.۲ The Predicate Calculus
۷.۲.۱ The Syntax of Predicates and Sentences
۷.۲.۲ A Semantics of Predicate Calculus
۷.۲.۳ A Blocks World Example of Semantic Meaning
۷.۳ Summary
۷.۴ Exercises
۸: The Predicate Calculus and Unification
۸.۱ Using Reasoning Rules to Produce New Predicate Expressions
۸.۲ The Unification Algorithm
۸.۲.۱ Unification: An Example
۸.۲.۲ Unification Application: A Logic-Based Financial Advisor
۸.۲.۳ The Financial Advisor: Using Heuristics
۸.۳ Summary
۸.۴ Exercises
۹: Resolution: Reasoning with the Propositional and Predicate Calculi
۹.۱ Introduction to Resolution
۹.۲ Resolution Refutation Systems
۹.۲.۱ Producing the Clause Form for Resolution Refutations
۹.۲.۲ The Binary Resolution Proof Procedure
۹.۳ Strategies and Simplification Techniques for Resolution
۹.۳.۱ The Breadth-First Strategy
۹.۳.۲ The Set of Support Strategy
۹.۳.۳ The Unit Preference Strategy
۹.۳.۴ The Linear Input Form Strategy
۹.۳.۵ Other Strategies and Techniques
۹.۴ Extracting Answers from Resolution Refutations
۹.۵ Logic Programming and Prolog
۹.۵.۱ Introduction
۹.۵.۲ Logic Foundations for Prolog
۹.۵.۳ Prolog and Automated Reasoning
۹.۶ Summary
۹.۷ Exercises
IV: Advanced Applications of Symbol-Based AI
۱۰: The Production System Representation and Search Engine
۱۰.۱ The Production System: An Architecture for Organizing Search
۱۰.۲ More Examples of Production System Problem-Solving
۱۰.۲.۱ The 8-Puzzle Revisited
۱۰.۲.۲ The Knight’s Tour Problem
۱۰.۳ The Expert System
۱۰.۴ The Physical Symbol System Hypothesis and the Birth of Cognitive Science
۱۰.۵ Summary
۱۰.۶ Exercises
۱۱: Advanced Applications of Symbol-Based AI: Planning and Learning
۱۱.۱ Introduction to Planning and Robotics
۱۱.۲ Using Planning Macros: STRIPS
۱۱.۳ Model-Based Planning: A NASA Example (Williams and Nayak 1996, 1997)
۱۱.۴ Symbol-Based Learning
۱۱.۴.۱ Decision Trees
۱۱.۴.۲ Reinforcement Learning
۱۱.۵ Bacon: Modeling the Celestial Environment
۱۱.۶ Expertise Wherever It Is Needed
۱۱.۷ Summary
۱۱.۸ Exercises
۱۲: Uncertain Reasoning: Symbol Based
۱۲.۱ Logic-Based Reasoning in Uncertain Situations
۱۲.۱.۱ Nonmonotonic Reasoning
۱۲.۱.۲ Truth Maintenance Systems
۱۲.۱.۳ Logics Based on Minimum Models
۱۲.۲ Uncertain Reasoning: Alternatives to Logic
۱۲.۲.۱ The Stanford Certainty Factor Algebra
۱۲.۲.۲ Reasoning with Fuzzy Sets
۱۲.۳ Summary and Pointers to Parts VI and VII
۱۲.۴ Exercises
V: Symbol-Based Associational Models for AI
۱۳: Introduction to Association-Based Knowledge Representations
۱۳.۱ The Behaviorist Tradition and Semantic Networks
۱۳.۱.۱ The Foundation for Graphical Representations of Meaning
۱۳.۱.۲ Semantic Networks
۱۳.۲ Conceptual Dependencies
۱۳.۳ Scripts
۱۳.۴ Summary
۱۳.۵ Exercises
۱۴: Association-Based Representations: Frames, Conceptual Graphs, WordNet, and FrameNet
۱۴.۱ Frames
۱۴.۲ Conceptual Graphs
۱۴.۲.۱ An Introduction to Conceptual Graphs
۱۴.۲.۲ Types, Individuals, and Names
۱۴.۲.۳ The Type Hierarchy
۱۴.۲.۴ Generalization and Specialization
۱۴.۲.۵ Propositional Nodes and Transformations to Logic
۱۴.۳ Current Uses of Association-Based Representations
۱۴.۴ Summary
۱۴.۵ Exercises
VI: Neural or Connectionist Networks
۱۵: An Introduction to Neural Networks
۱۵.۱ An Artificial Neuron and Applications
۱۵.۲ Early Research: McCulloch, Pitts, and Hebb
۱۵.۳ Perceptrons
۱۵.۳.۱ An Example: Perceptron Classification
۱۵.۳.۲ Linear Separability and the Exclusive-or Problem
۱۵.۴ Summary
۱۵.۵ Exercises
۱۶: The Delta Rule, Backpropagation, and Matrix Representations
۱۶.۱ The Generalized Delta Rule
۱۶.۲ The Backpropagation Algorithm
۱۶.۲.۱ Example: Backpropagation Solving the Exclusive-Or Problem
۱۶.۳ Matrix Representations for Network Processing
۱۶.۳.۱ A Review of Matrix Algebra
۱۶.۳.۲ A Review of Vector Algebra and the Cosine Distance Calculation
۱۶.۴ Matrix Representations and Neural Network Solutions
۱۶.۴.۱ Representing Perceptron Networks
۱۶.۴.۲ Representing Feed-Forward Backpropagation Networks
۱۶.۴.۳ Meta-parameters for Small Networks
۱۶.۵ Summary
۱۶.۶ Exercises
۱۷: Deep Learning: Introduction and Representations
۱۷.۱ Toward Deep Learning
۱۷.۱.۱ From Backpropagation to Deep Learning
۱۷.۱.۲ AlphaGo Zero and Alpha Zero
۱۷.۱.۳ Robot Navigation: PRM-RL
۱۷.۱.۴ Deep Learning and Video Games
۱۷.۱.۵ Deep Learning and Natural Language Understanding
۱۷.۲ Meta-parameters for Very Large Networks
۱۷.۲.۱ Softmax: Representing Network Output as a Probability Distribution
۱۷.۲.۲ Softmax Activation and Cross-Entropy Loss
۱۷.۲.۳ Data Handling in Network Learning
۱۷.۳ Alternative Architectures: Convolutional and Recurrent Networks
۱۷.۳.۱ Convolutional Neural Networks
۱۷.۳.۲ Recurrent Neural Networks and Long Short-Term Memory
۱۷.۴ Autoencoders and Transfer Learning
۱۷.۴.۱ Autoencoders
۱۷.۴.۲ Transfer Learning
۱۷.۴.۲.۱ Algorithms, and Related Data Structures
۱۷.۴.۲.۲ Transfer Learning with Latent Semantic Analysis and Transformers
۱۷.۵ Summary
۱۷.۶ Exercises
۱۸: Building Language Models and Transformers
۱۸.۱ Latent Semantic Analysis: Distributed Semantic Representations
۱۸.۲ Building Large Language Models
۱۸.۲.۱ Preparing the Text
۱۸.۲.۲ Creating Word/Token Embeddings
۱۸.۲.۳ Bigrams and Trigrams
۱۸.۲.۴ Training the Language Model
۱۸.۳ Toward Transformer-Based Large Language Models
۱۸.۳.۱ Transformer Models
۱۸.۳.۲ Attention
۱۸.۴ The Transformer in Practice
۱۸.۴.۱ Pretraining: The Corpora
۱۸.۴.۲ Fine-Tuning the LLM
۱۸.۴.۳ Prompt Engineering
۱۸.۴.۴ Sample Applications of Generative AI
۱۸.۴.۴.۱ Image- to-Image Conversion (url: 18.1)
۱۸.۴.۴.۲ Text to Images and Images to Text (url: 18.4 and url: 18.10)
۱۸.۴.۴.۳ Text to Video
۱۸.۴.۴.۴ Music Production (url: 18.6)
۱۸.۴.۴.۵ Code Generation (url: 18.2 and url: 18.12)
۱۸.۴.۴.۶ Protein Design and Generation (url: 18.13)
۱۸.۴.۴.۷ Continued Extensions of Google’s DeepMind Problem Solvers (url: 18.14)
۱۸.۵ Summary
۱۸.۶ Exercises
۱۹: Alternative Network Architectures: Prototypes and Classifiers
۱۹.۱ A Kohonen Network: Winner-Takes-All Classification
۱۹.۲ A Kohonen Network: Learning Prototypes
۱۹.۳ Outstar Networks and Counterpropagation
۱۹.۴ Supervised Hebbian Learning
۱۹.۵ Associative Memories and the Linear Associator
۱۹.۶ Summary
۱۹.۷ Exercises
۲۰: Alternative Network Architectures: Attractor Networks and Memories
۲۰.۱ Introduction to Associative Memories
۲۰.۲ BAM, the Bidirectional Associative Memory
۲۰.۲.۱ The BAM Network Architecture
۲۰.۲.۲ Examples of BAM Processing
۲۰.۳ Autoassociative Memories and Hopfield Networks
۲۰.۴ Summary
۲۰.۵ Exercises
VII: Probabilistic Artificial Intelligence
۲۱: Counting, the Foundation for Probabilities
۲۱.۱ Introduction to Probabilistic Reasoning
۲۱.۲ The Elements of Counting
۲۱.۲.۱ The Addition and Multiplication Rules
۲۱.۲.۲ Permutations and Combinations
۲۱.۳ Elements of Probability Theory
۲۱.۳.۱ The Sample Space, Probabilities, and Independence
۲۱.۳.۲ Probabilistic Inference: An Example
۲۱.۴ Summary
۲۱.۵ Exercises
۲۲: Bayes’ Theorem
۲۲.۱ Random Variables
۲۲.۲ Conditional Probability and an Introduction to Bayesian Reasoning
۲۲.۳ Bayes’ Theorem
۲۲.۴ Two Examples of Bayesian Reasoning
۲۲.۴.۱ Example: Purchasing an Automobile
۲۲.۴.۲ Example: Extending the Traffic Slowdown Problem
۲۲.۵ Summary
۲۲.۶ Exercises
۲۳: Bayesian Belief Networks and Observable Markov Models
۲۳.۱ Introduction to Stochastic Models
۲۳.۲ A Directed Graphical Model: The Bayesian Belief Network
۲۳.۲.۱ Directed Graphical Models: d-separation
۲۳.۲.۲ Graphical Models: An Inference Algorithm
۲۳.۳ Dynamic Bayesian Networks
۲۳.۴ Observable Markov Models
۲۳.۵ Summary
۲۳.۶ Exercises
۲۴: Hidden Markov and Alternative Probabilistic Models
۲۴.۱ Hidden Markov Models
۲۴.۲ Important Variants of Hidden Markov Models
۲۴.۲.۱ Auto-Regressive HMMs, AR-HMMs
۲۴.۲.۲ Factorial HMMs
۲۴.۲.۳ N-Gram HMMs
۲۴.۳ A Short Survey of Alternative Markov Models
۲۴.۳.۱ Hierarchical HMMs, HHMMs
۲۴.۳.۲ Semi-Markov Models
۲۴.۳.۳ Markov Decision Processes
۲۴.۴ First-Order Alternatives to BBNs and HMMs
۲۴.۴.۱ Bayesian Network Construction from Knowledge Bases
۲۴.۴.۲ Bayesian Logic Programs, BLPs
۲۴.۴.۳ Probabilistic Relational Models, PRMs
۲۴.۴.۴ Markov Logic Networks, MLNs
۲۴.۴.۵ Loopy Logic
۲۴.۵ Two Stochastic Engineering Examples: An MDP and a DBN
۲۴.۵.۱ Using a Markov Decision Process to Model a Reward-Driven Robot
۲۴.۵.۲ A Dynamic Bayesian Network Model of Nuclear Power Generation
۲۴.۶ Summary
۲۴.۷ Exercises
VIII: AI: Ethical Issues, Fundamental Limitations, and Future Promise
۲۵: Artificial Intelligence: User’s Ethical Issues
۲۵.۱ Artificial and Human Intelligence
۲۵.۱.۱ The Foundations of Artificial Intelligence
۲۵.۱.۲ Artificial and Human Intelligence: A Category Difference with Ethical Consequences
۲۵.۲ Ethical Issues from the Perspective of the AI User
۲۵.۲.۱ Artificial Intelligence vs Human Intelligence
۲۵.۲.۲ AI Products Are Human Created and Require Human Responsibility
۲۵.۲.۳ AI Computation Has an Environmental Cost
۲۵.۳ We Are Already Embedded in an AI World
۲۵.۳.۱ How to I Perform Responsible Internet Research?
۲۵.۳.۲ Who Owns AI-Generated Intellectual Property?
۲۵.۳.۳ Who Is the Merchandise in an Internet Environment?
۲۵.۳.۴ Who Owns Your Persona?
۲۵.۳.۵ An Ethical Stance Requires Being Aware of AI’s Limitations
۲۵.۴ Summary
۲۵.۵ Exercises
۲۶: AI Ethical Issues: From a Social Perspective
۲۶.۱ Building Society-Oriented AI Projects
۲۶.۲ If We Do Build These AI Projects, What Levels of Protection Are Needed?
۲۶.۲.۱ Ethical Issues for AI Designers and Programmers
۲۶.۲.۲ Ethical Issues for AI Sales and Management
۲۶.۲.۳ Ethical Issues for the Purchaser of AI Products
۲۶.۳ What Is Data?
۲۶.۴ Data and Algorithm Transparency and Explainability
۲۶.۵ Design Honesty, Transparency, Explanations, Data Bias, and Privacy
۲۶.۵.۱ Design Knowledge and Honest Technical Presentation
۲۶.۵.۲ Transparency and Explanations
۲۶.۵.۳ Medical Diagnostics, Explanations, and Privacy
۲۶.۵.۴ Natural Language and Sentiment Analysis
۲۶.۵.۵ Dealing with Dynamic Systems
۲۶.۵.۶ Privacy: The User as Merchandise
۲۶.۵.۷ Protection of Intellectual Property
۲۶.۶ Safeguarding AI Research and Practice
۲۶.۷ Summary
۲۶.۸ Exercises
۲۷: AI: Philosophical Perspectives, Current Limitations, and Future Promise
۲۷.۱ Modern AI: A Psychological, Mathematical, and Philosophical Perspective
۲۷.۲ Several Fundamental Limitations of Current AI Technology
۲۷.۲.۱ Limitations of Current AI Practice
۲۷.۲.۲ What Is the Role of Embodiment and Culture in Intelligence?
۲۷.۲.۳ Meaning: The Grounding Problem
۲۷.۳ Artificial Intelligence: Where Are We Going?
Bibliography
URL
Index