Preface
1 Introduction
 1.1 WELCOME TO THE NEW PDP HANDBOOK
 1.2 MODELS, PROGRAMS, CHAPTERS AND EXCERCISES
  1.2.1 Key Features of PDP Models
 1.3 SOME GENERAL CONVENTIONS AND CONSIDERATIONS
  1.3.1 Mathematical Notation
  1.3.2 Pseudo-MATLAB Code
  1.3.3 Program Design and User Interface
  1.3.4 Exploiting the MATLAB Envioronment
 1.4 BEFORE YOU START
 1.5 MATLAB MINI-TUTORIAL
  1.5.1 Basic Operations
  1.5.2 Vector Operations
  1.5.3 Logical operations
  1.5.4 Control Flow
  1.5.5 Vectorized Code
2 Interactive Activation and Competition
 2.1 BACKGROUND
  2.1.1 How Competition Works
  2.1.2 Resonance
  2.1.3 Hysteresis and Blocking
  2.1.4 Grossberg’s Analysis of Interactive Activation and Competition Processes
 2.2 THE IAC MODEL
  2.2.1 Architecture
  2.2.2 Visible and Hidden Units
  2.2.3 Activation Dynamics
  2.2.4 Parameters
  2.2.5 Pools and Projections
  2.2.6 The Core Routines
 2.3 EXERCISES
3 Constraint Satisfaction in PDP Systems
 3.1 BACKGROUND
 3.2 THE SCHEMA MODEL
 3.3 IMPLEMENTATION
 3.4 RUNNING THE PROGRAM
  3.4.1 Reset, Newstart, and the Random Seed
  3.4.2 Options and parameters
 3.5 OVERVIEW OF EXERCISES
 3.6 GOODNESS AND PROBABILITY
  3.6.1 Local Maxima
  3.6.2 Escaping from Local Maxima
4 Learning in PDP Models: The Pattern Associator
 4.1 BACKGROUND
  4.1.1 The Hebb Rule
  4.1.2 The Delta Rule
  4.1.3 Division of Labor in Error Correcting Learning
  4.1.4 The Linear Predictability Constraint
 4.2 THE PATTERN ASSOCIATOR
  4.2.1 The Hebb Rule in Pattern Associator Models
  4.2.2 The Delta Rule in Pattern Associator Models
  4.2.3 Linear Predictability and the Linear Independence Requirement
  4.2.4 Nonlinear Pattern Associators
 4.3 THE FAMILY OF PATTERN ASSOCIATOR MODELS
  4.3.1 Activation Functions
  4.3.2 Learning Assumptions
  4.3.3 The Environment and the Training Epoch
  4.3.4 Performance Measures
 4.4 IMPLEMENTATION
 4.5 RUNNING THE PROGRAM
  4.5.1 Commands and Parameters
  4.5.2 State Variables
 4.6 OVERVIEW OF EXERCISES
  4.6.1 Further Suggestions for Exercises
5 Training Hidden Units with Back Propagation
 5.1 BACKGROUND
  5.1.1 Minimizing Mean Squared Error
  5.1.2 The Back Propagation Rule
 5.2 IMPLEMENTATION
 5.3 RUNNING THE PROGRAM
 5.4 EXERCISES
6 Competitive Learning
 6.1 SIMPLE COMPETITIVE LEARNING
  6.1.1 Background
  6.1.2 Some Features of Competitive Learning
  6.1.3 Implementation
  6.1.4 Overview of Exercises
 6.2 SELF-ORGANIZING MAP
  6.2.1 The Model
  6.2.2 Some Features of the SOM
  6.2.3 Implementation
  6.2.4 Overview of Exercises
7 The Simple Recurrent Network: A Simple Model that Captures the Structure in Sequences
 7.1 BACKGROUND
  7.1.1 The Simple Recurrent Network
  7.1.2 Graded State Machines
 7.2 THE SRN PROGRAM
  7.2.1 Sequences
  7.2.2 New Parameters
  7.2.3 Network specification
  7.2.4 Fast mode for training
 7.3 EXERCISES
8 Recurrent Backpropagation: Attractor network models of semantic and lexical processing
 8.1 BACKGROUND
 8.2 THE RBP PROGRAM
  8.2.1 Time intervals, and the partitioning of intervals into ticks
  8.2.2 Visualizing the state space of an rbp network
  8.2.3 Forward propagation of activation.
  8.2.4 Backward propagation of error
  8.2.5 Calculating the weight error derivatives
  8.2.6 Updating the weights.
 8.3 Using the rbp program with the rogers network
  8.3.1 rbp fast training mode.
  8.3.2 Training and Lesioning with the rogers network
  8.3.3 rbp pattern files.
  8.3.4 Creating an rbp network
9 Temporal-Difference Learning
 9.1 BACKGROUND
 9.2 REINFORCEMENT LEARNING
  9.2.1 Discounted Returns
  9.2.2 The Control Problem
 9.3 TD AND BACK PROPAGATION
  9.3.1 Back Propagating TD Error
  9.3.2 Case Study: TD-Gammon
 9.4 IMPLEMENTATION
  9.4.1 Specifying the Environment
 9.5 RUNNING THE PROGRAM
 9.6 EXERCISES
A PDPTool Installation and Quick Start Guide
 A.1 System requirements
 A.2 Installation
 A.3 Using PDPTool at a Stanford Cluster Computer
 A.4 Using the software
 A.5 Notes when using Matlab 7.3 r2006b on OSX
B How to create your own network
 B.1 Creating the network itself
  B.1.1 Defining the Network Pools
  B.1.2 Defining the Projections
 B.2 Creating the display template
 B.3 Creating the example file
 B.4 Creating a script to initialize the network
C PDPTool User’s Guide
D PDPTool Standalone Executable
 D.1 Installing under Linux
 D.2 Running under Linux
 D.3 Installing under Windows
 D.4 Running under Windows
 D.5 Installing under Mac OSX
 D.6 Running under Mac OSX