Computational Models for Neuroscience

The cerebral cortex is the area of the brain where almost all of the mental capabilities that distinguish humans from other animals are based. Understanding how it works is vital to the progress in many areas of medicine and science, including the development of more efficient and flexible neural networks. Although extensive research has been carried out over the last 150 years, very little real progress has been made. This book provides a unique overview of work by key researchers, and shows how current research can be used to develop and improve computational models for neural networks.

1 The Neurointeractive Paradigm: Dynamical Mechanics and the Emergence of Higher Cortical Function.- 1.1 Abstract.- 1.2 Introduction.- 1.3 Principles of Cortical Neurointeractivity.- 1.4 Dynamical Mechanics.- 1.5 The Neurointeractive Cycle.- 1.6 Developmental Emergence.- 1.7 Explaining Emergence.- 1.8 References.- 2 The Cortical Pyramidal Cell as a Set of Interacting Error Backpropagating Dendrites: Mechanism for Discovering Nature ’s Order.- 2.1 Abstract.- 2.2 Introduction.- 2.2.1 Defining the Problem.- 2.2.2 How Does the Brain Discover Orderly Relations?.- 2.3 Implementation of the Proposal.- 2.3.1 How Might Error Backpropagation Learning Be Implemented in Dendrites?.- 2.3.2 How Can Dendrites Be Set Up to Teach Each Other?.- 2.3.3 How to Divide Connections Among the Dendrites?.- 2.4 Cortical Minicolumnar Organization and SINBAD Neurons.- 2.5 Associationism.- 2.5.1 SINBAD as an Associationist Theory.- 2.5.2 Countering Nativist Arguments.- 2.6 Acknowledgements.- References.- 3 Performance of Intelligent Systems Governed by Internally Generated Goals.- 3.1 Abstract.- 3.2 Introduction.- 3.3 Perception as an Active Process.- 3.4 Nonlinear Dynamics of the Olfactory System.- 3.5 Chaotic Oscillations During Learning Novel Stimuli.- 3.6 Generalization and Consolidation of New Perceptions with Context.- 3.7 The Central Role of the Limbic System.- 3.8 Conclusions.- 3.9 Acknowledgements.- References.- 4 A Theory of Thalamocortex.- 4.1 Abstract.- 4.2 Active Neurons.- 4.3 Neuronal Connections within Thalamocortex.- 4.4 Cortical Regions.- 4.5 Feature Artractor Associative Memory Neural Network.- 4.6 Antecedent Support Associative Memory Neural Network.- 4.7 Hierarchical Abstractor Associative Memory Neural Network.- 4.8 Consensus Building.- 4.9 Brain Command Loop.- 4.10 Testing this Theory.- 4.11 Acknowledgements.- Appendix A: Sketch of an Analysis of the Simplified Feature Artractor Associative Memory Neural Network.- Appendix B: Experiments with a Simplified Antecedent Support Associative Memory Neural Network.- Appendix C: An Experiment with Consensus Building.- References.- 5 Elementary Principles of Nonlinear Synaptic Transmission.- 5.1 Abstract.- 5.2 Introduction.- 5.3 Frequency-dependent Synaptic Transmission.- 5.4 Nonlinear Synapses Enable Temporal Integration.- 5.5 Temporal Information.- 5.6 Packaging Temporal Information.- 5.7 Size of Temporal Information Packages.- 5.8 Classes of Temporal Information Packages.- 5.9 Emergence of the Population Signal.- 5.10 Recurrent Neural Networks.- 5.11 Combining Temporal Information in Recurrent Networks.- 5.12 Organization of Synaptic Parameters.- 5.13 Learning Dynamics, Learning to Predict.- 5.14 Redistribution of Synaptic Efficacy.- 5.15 Optimizing Synaptic Prediction.- 5.16 A Nested Learning Algorithm.- 5.17 Retrieving Memories from Nonlinear Synapses.- 5.18 Conclusion.- 5.19 Acknowledgements.- Appendix A: Sherrington ’s Leap.- Appendix B: Functional Significance.- Appendix C: Visual Patch Recordings.- Appendix D: Biophysical Basis of Parameters.- Appendix E: Single Connection, Many Synapses.- Appendix F: The Model.- Appendix G: Synaptic Classes.- Appendix H: Paired Pulses.- Appendix I: Digitization of Synaptic Parameters.- Appendix J: Steady State.- Appendix K: Inhibitory Synapses.- Appendix L: Lack of Boundaries.- Appendix M: Speed of RI Accumulation.- Appendix N: Network Efficiency.- Appendix O: The Binding Problem of the Binding Problem.- References.- 6 The Development of Cortical Models to Enable Neural-based Cognitive Architectures.- 6.1 Introduction.- 6.1.1 Computational Neuroscience Paradigms and Predictions.- 6.2 The Challenge of Cognitive Architectures.- 6.2.1 General Cognitive Skills.- 6.2.2 A Survey of Current Cognitive Architectures.- 6.2.3 Assumptions and Limitations of Current Cognitive Architectures.- 6.3 The Prospects for a Neural-based Cognitive Architecture.- 6.3.1 Limitations of Artificial Neural Networks.- 6.3.2 Biological Networks Emerging from Computational Neuroscience: Sensory and Motor Modules.- 6.3.3 Forebrain Systems Supporting Cortical Function.- 6.4 Elements of a General Cortical Model.- 6.4.1 Single Neuron Models or Processor Elements.- 6.4.2 Microcircuitry.- 6.4.3 Dynamic Synaptic Connectivity.- 6.4.4 Ensemble Dynamics and Coding.- 6.4.5 Transient Coherent Structures and Cognitive Dynamics.- 6.5 Promising Models and their Capabilities.- 6.5.1 Biologically Based Cortical Systems.- 6.5.2 A Cortical System Based on Neurobiology, Biological Principles and Mathematical Analysis: Cortronics.- 6.5.3 Connectionist Architectures with Biological Principles: The Convergence of Cognitive Science and Computational Neuroscience.- 6.6 The Challenges of Demonstrating Cognitive Ability.- 6.6.1 Robotics and Autonomous Systems.- 6.7 Co-development Strategies for Automated Systems and Human Performers.- 6.8 Acknowledgements.- References.- 7 The Behaving Human Neocortex as a Dynamic Network of Networks.- 7.1 Abstract.- 7.2 Neural Organization Across Scales.- 7.3 Network of Networks (NoN) Model.- 7.3.1 Architecture.- 7.3.2 Model Formulation.- 7.3.3 NoN Properties.- 7.3.4 NoN Contributions.- 7.4 Neurobiological Predicatability and Falsifiability.- 7.5 Implications for Neuroengineering.- 7.6 Concluding Remarks.- 7.7 Acknowledgements.- References.- 8 Towards Global Principles of Brain Processing.- 8.1 Abstract.- 8.2 Introduction.- 8.3 What Could Brain Principles Look Like?.- 8.4 Structural Modeling.- 8.5 Static Activation Study Results.- 8.6 The Motion After-Effect (MAE).- 8.7 The Three-Stage Model of Consciousness.- 8.8 The CODAM Model of Consciousness.- 8.9 Principles of the Global Brain.- 8.10 The Thinking Brain.- 8.11 Discussion.- 8.12 Acknowledgement.- References.- 9 The Neural Networks for Language in the Brain: Creating LAD.- 9.1 Abstract.- 9.2 Introduction.- 9.3 The ACTION Net Model of TSSG.- 9.4 Phrase Structure Analyzers.- 9.5 Generativity of the Adjectival Phrase Analyzer.- 9.6 Complexity of Phrase Structure Analysis.- 9.7 Future Directions in the Construction of LAD.- 9.8 Conclusions.- References.- 10 Cortical Belief Networks.- 10.1 Abstract.- 10.1 Introduction.- 10.1 An Example.- 10.1 Representing Distributions in Populations.- 10.1 Basis Function Representations.- 10.1 Generative Representations.- 10.1 Standard Bayesian Approach.- 10.1 Distributional Population Coding.- 10.1 Applying Distributional Population Coding.- 10.1.1 Population Analysis.- 10.1.1 Decoding Transparent Motion.- 10.1.1 Decision Noise.- 10.1.1 Lateral Interactions.- 10.1 Cortical Belief Network.- 10.1 Discussion.- 10.1 Acknowledgements.- 10.1 References.