Atjaunināt sīkdatņu piekrišanu

Cerebral Cortex: Principles of Operation [Hardback]

4.67/5 (12 ratings by Goodreads)
(Oxford Centre for Computational Neuroscience, Oxford, UK)
  • Formāts: Hardback, 992 pages, height x width x depth: 247x179x42 mm, weight: 1564 g
  • Izdošanas datums: 18-Aug-2016
  • Izdevniecība: Oxford University Press
  • ISBN-10: 0198784856
  • ISBN-13: 9780198784852
Citas grāmatas par šo tēmu:
Cerebral Cortex: Principles of OperationPalielināt
  • Formāts: Hardback, 992 pages, height x width x depth: 247x179x42 mm, weight: 1564 g
  • Izdošanas datums: 18-Aug-2016
  • Izdevniecība: Oxford University Press
  • ISBN-10: 0198784856
  • ISBN-13: 9780198784852
Citas grāmatas par šo tēmu:
Permanent link: https://www.kriso.lv/db/9780198784852.html
Keywords:
"A book remarkable in its ambition, and even more remarkable in its content. A truly landmark achievement by a neuroscientist who has brought together his lifetime of research knowledge and experience into this outstanding volume. Edmund Rolls is to be congratulated on this impressive synthesis of decades of neuroscience data."

David Nutt, Professor of Neuropsychopharmacology at Imperial College London and President of the European Brain Council

The aim of this book is to provide insight into the principles of operation of the cerebral cortex. These principles are key to understanding how we, as humans, function.

There have been few previous attempts to set out some of the important principles of operation of the cortex, and this book is pioneering. The book goes beyond separate connectional neuroanatomical, neurophysiological, neuroimaging, neuropsychiatric, and computational neuroscience approaches, by combining evidence from all these areas to formulate hypotheses about how and what the cerebral cortex computes. As clear hypotheses are needed in this most important area of 21st century science, how our brains work, the author has formulated a set of hypotheses about the principles of cortical operation to guide thinking and future research.

The book focusses on the principles of operation of the cerebral cortex, because at this time it is possible to propose and describe many principles, and many are likely to stand the test of time, and provide a foundation for further developments, even if some need to be changed. In this context, I have not attempted to produce an overall theory of operation of the cerebral cortex, because at this stage of our understanding, such a theory would be incorrect or incomplete. However, many of the principles described will provide the foundations for more complete theories of the operation of the cerebral cortex. This book is intended to provide a foundation for future understanding, and it is hoped that future work will develop and add to these principles of operation of the cerebral cortex.

The book includes Appendices on the operation of many of the neuronal networks described in the book, together with simulation software written in Matlab.

This book will be valuable to all those interested in understanding our cerebral cortex and how it operates to account for many aspects of brain function and cognitive function in health and disease. The book is relevant to those in the areas of neuroscience, neurology, psychology, psychiatry, computational neuroscience, biology, and philosophy.



Professor Edmund T. Rolls performs full-time research at the Oxford Centre for Computational Neuroscience, and is professor of Computational Neuroscience at the University of Warwick, and has acted as Professor of Experimental Psychology at the University of Oxford, and as Fellow and Tutor of Corpus Christi College, Oxford. His research links neurophysiological and computational neuroscience approaches to human functional neuroimaging and neuropsychological studies in order to provide a fundamental basis for understanding human brain function and its disorders.
1 Introduction 1(39)
1.1 Principles of operation of the cerebral cortex: introduction and plan
1(3)
1.2 Neurons
4(2)
1.3 Neurons in a network
6(2)
1.4 Synaptic modification
8(1)
1.5 Long-term potentiation and long-term depression
9(5)
1.6 Distributed representations
14(2)
1.6.1 Definitions
14(1)
1.6.2 Advantages of different types of coding
15(1)
1.7 Neuronal network approaches versus connectionism
16(1)
1.8 Introduction to three neuronal network architectures
17(1)
1.9 Systems-level analysis of brain function
18(9)
1.9.1 Ventral cortical visual stream
19(2)
1.9.2 Dorsal cortical visual stream
21(2)
1.9.3 Hippocampal memory system
23(1)
1.9.4 Frontal lobe systems
23(1)
1.9.5 Brodmann areas
24(3)
1.10 The fine structure of the cerebral neocortex
27(12)
1.10.1 The fine structure and connectivity of the neocortex
27(1)
1.10.2 Excitatory cells and connections
27(2)
1.10.3 Inhibitory cells and connections
29(3)
1.10.4 Quantitative aspects of cortical architecture
32(2)
1.10.5 Functional pathways through the cortical layers
34(4)
1.10.6 The scale of lateral excitatory and inhibitory effects, and modules
38(1)
1.11 Highlights
39(1)
2 Hierarchical organization 40(32)
2.1 Introduction
40(1)
2.2 Hierarchical organization in sensory systems
41(26)
2.2.1 Hierarchical organization in the ventral visual system
41(5)
2.2.2 Hierarchical organization in the dorsal visual system
46(2)
2.2.3 Hierarchical organization of taste processing
48(9)
2.2.4 Hierarchical organization of olfactory processing
57(2)
2.2.5 Hierarchical multimodal convergence of taste, olfaction, and vision
59(5)
2.2.6 Hierarchical organization of auditory processing
64(3)
2.3 Hierarchical organization of reward value processing
67(1)
2.4 Hierarchical organization of connections to the frontal lobe for short-term memory
68(1)
2.5 Highlights
69(3)
3 Localization of function 72(3)
3.1 Hierarchical processing
72(1)
3.2 Short-range neocortical recurrent collaterals
72(1)
3.3 Topographic maps
72(1)
3.4 Modularity
72(1)
3.5 Lateralization of function
73(1)
3.6 Ventral and dorsal cortical areas
73(1)
3.7 Highlights
74(1)
4 Recurrent collateral connections and attractor networks 75(16)
4.1 Introduction
75(1)
4.2 Attractor networks implemented by the recurrent collaterals
75(1)
4.3 Evidence for attractor networks implemented by recurrent collateral connections
76(4)
4.3.1 Short-term Memory
77(3)
4.3.2 Long-term Memory
80(1)
4.3.3 Decision-Making
80(1)
4.4 The storage capacity of attractor networks
80(1)
4.5 A global attractor network in hippocampal CA3, but local in neocortex
81(2)
4.6 The speed of operation of cortical attractor networks
83(1)
4.7 Dilution of recurrent collateral cortical connectivity
83(2)
4.8 Self-organizing topographic maps in the neocortex
85(1)
4.9 Attractors formed by forward and backward connections between cortical areas?
85(1)
4.10 Interacting attractor networks
86(4)
4.11 Highlights
90(1)
5 The noisy cortex: stochastic dynamics, decisions, and memory 91(50)
5.1 Reasons why the brain is inherently noisy and stochastic
91(4)
5.2 Attractor networks, energy landscapes, and stochastic neurodynamics
95(3)
5.3 A multistable system with noise
98(3)
5.4 Stochastic dynamics and the stability of short-term memory
101(5)
5.4.1 Analysis of the stability of short-term memory
103(1)
5.4.2 Stability and noise in a model of short-term memory
104(2)
5.5 Long-term memory recall
106(1)
5.6 Stochastic dynamics and probabilistic decision-making in an attractor network
106(28)
5.6.1 Decision-making in an attractor network
107(1)
5.6.2 Theoretical framework: a probabilistic attractor network
107(3)
5.6.3 Stationary multistability analysis: mean-field
110(2)
5.6.4 Integrate-and-fire simulations of decision-making: spiking dynamics
112(4)
5.6.5 Reaction times of the neuronal responses
116(1)
5.6.6 Percentage correct
117(1)
5.6.7 Finite-size noise effects
117(2)
5.6.8 Comparison with neuronal data during decision-making
119(3)
5.6.9 Testing the model of decision-making with human functional neuroimaging
122(7)
5.6.10 Decisions based on confidence in one's decisions: self-monitoring
129(2)
5.6.11 Decision-making with multiple alternatives
131(1)
5.6.12 The matching law
132(1)
5.6.13 Comparison with other models of decision-making
132(2)
5.7 Perceptual decision-making and rivalry
134(1)
5.8 Symmetry-breaking
135(1)
5.9 The evolutionary utility of probabilistic choice
135(1)
5.10 Selection between conscious vs unconscious decision-making, and free will
136(1)
5.11 Creative thought
137(1)
5.12 Unpredictable behaviour
138(1)
5.13 Predicting a decision before the evidence is applied
138(2)
5.14 Highlights
140(1)
6 Attention, short-term memory, and biased competition 141(45)
6.1 Bottom-up attention
141(2)
6.2 Top-down attention - biased competition
143(28)
6.2.1 The biased competition hypothesis
143(2)
6.2.2 Biased competition - single neuron studies
145(2)
6.2.3 Non-spatial attention
147(2)
6.2.4 Biased competition - fMRI
149(1)
6.2.5 A basic computational module for biased competition
149(1)
6.2.6 Architecture of a model of attention
150(4)
6.2.7 Simulations of basic experimental findings
154(4)
6.2.8 Object recognition and spatial search
158(5)
6.2.9 The neuronal and biophysical mechanisms of attention
163(4)
6.2.10 'Serial' vs 'parallel' attentional processing
167(4)
6.3 Top-down attention - biased activation
171(10)
6.3.1 Selective attention can selectively activate different cortical areas
171(2)
6.3.2 Sources of the top-down modulation of attention
173(1)
6.3.3 Granger causality used to investigate the source of the top-down biasing
174(1)
6.3.4 Top-down cognitive modulation
175(3)
6.3.5 A top-down biased activation model of attention
178(3)
6.4 Conclusions
181(3)
6.5 Highlights
184(2)
7 Diluted connectivity 186(23)
7.1 Introduction
186(1)
7.2 Diluted connectivity and the storage capacity of attractor networks
187(11)
7.2.1 The autoassociative or attractor network architecture being studied
187(1)
7.2.2 The storage capacity of attractor networks with diluted connectivity
188(2)
7.2.3 The network simulated
190(2)
7.2.4 The effects of diluted connectivity on the capacity of attractor networks
192(5)
7.2.5 Synthesis of the effects of diluted connectivity in attractor networks
197(1)
7.3 The effects of dilution on the capacity of pattern association networks
198(3)
7.4 The effects of dilution on the performance of competitive networks
201(6)
7.4.1 Competitive Networks
201(1)
7.4.2 Competitive networks without learning but with diluted connectivity
202(1)
7.4.3 Competitive networks with learning and with diluted connectivity
203(2)
7.4.4 Competitive networks with learning and with full (undiluted) connectivity
205(1)
7.4.5 Overview and implications of diluted connectivity in competitive networks
206(1)
7.5 The effects of dilution on the noise in attractor networks
207(1)
7.6 Highlights
207(2)
8 Coding principles 209(18)
8.1 Types of encoding
209(1)
8.2 Place coding with sparse distributed firing rate representations
210(11)
8.2.1 Reading the code used by single neurons
210(4)
8.2.2 Understanding the code provided by populations of neurons
214(7)
8.3 Synchrony, coherence, and binding
221(1)
8.4 Principles by which the representations are formed
222(1)
8.5 Information encoding in the human cortex
223(3)
8.6 Highlights
226(1)
9 Synaptic modification for learning 227(14)
9.1 Introduction
227(1)
9.2 Associative synaptic modification implemented by long-term potentiation
227(1)
9.3 Forgetting in associative neural networks, and memory reconsolidation
228(5)
9.3.1 Forgetting
228(2)
9.3.2 Factors that influence synaptic modification
230(2)
9.3.3 Recortsolidation
232(1)
9.4 Spike-timing dependent plasticity
233(1)
9.5 Long-term synaptic depression in the cerebellar cortex
233(1)
9.6 Reward prediction error learning
234(6)
9.6.1 Blocking and delta-rule learning
234(1)
9.6.2 Dopamine neuron firing and reward prediction error learning
234(6)
9.7 Highlights
240(1)
10 Synaptic and neuronal adaptation and facilitation 241(14)
10.1 Mechanisms for neuronal adaptation and synaptic depression and facilitation
241(3)
10.1.1 Sodium inactivation leading to neuronal spike-frequency adaptation
241(1)
10.1.2 Calcium activated hyper-polarizing potassium current
242(1)
10.1.3 Short-term synaptic depression and facilitation
243(1)
10.2 Short-term depression of thalamic input to the cortex
244(1)
10.3 Relatively little adaptation in primate cortex when it is operating normally
244(3)
10.4 Acetylcholine, noradrenaline, and other modulators of adaptation and facilitation
247(2)
10.4.1 Acetylcholine
247(1)
10.4.2 Noradrenergic neurons
248(1)
10.5 Synaptic depression and sensory-specific satiety
249(1)
10.6 Neuronal and synaptic adaptation, and the memory for sequential order
250(1)
10.7 Destabilization of short-term memory by adaptation or synaptic depression
250(1)
10.8 Non-reward computation in the orbitofrontal cortex using synaptic depression
251(2)
10.9 Synaptic facilitation and a multiple-item short-term memory
253(1)
10.10 Synaptic facilitation in decision-making
253(1)
10.11 Highlights
254(1)
11 Backprojections In the neocortex 255(7)
11.1 Architecture
255(2)
11.2 Learning
257(1)
11.3 Recall
258(1)
11.4 Semantic priming
259(1)
11.5 Top-down Attention
259(2)
11.6 Autoassociative storage, and constraint satisfaction
261(1)
11.7 Highlights
261(1)
12 Memory and the hippocampus 262(7)
12.1 Introduction
262(1)
12.2 Hippocampal circuitry and connections
262(1)
12.3 The hippocampus and episodic memory
262(1)
12.4 Autoassociation in the CA3 network for episodic memory
263(2)
12.5 The dentate gyrus as a pattern separation mechanism, and neurogenesis
265(1)
12.6 Rodent place cells vs primate spatial view cells
265(1)
12.7 Backprojections, and the recall of information from the hippocampus to neocortex
266(1)
12.8 Subcortical structures connected to the hippocampo-cortical memory system
267(1)
12.9 Highlights
267(2)
13 Limited neurogenesis in the adult cortex 269(3)
13.1 No neurogenesis in the adult neocortex
269(1)
13.2 Limited neurogenesis in the adult hippocampal dentate gyrus
269(1)
13.3 Neurogenesis in the chemosensing receptor systems
270(1)
13.4 Highlights
271(1)
14 Invariance learning and vision 272(9)
14.1 Hierarchical cortical organization with convergence
272(1)
14.2 Feature combinations
272(1)
14.3 Sparse distributed representations
273(1)
14.4 Self-organization by feedforward processing without a teacher
273(1)
14.5 Learning guided by the statistics of the visual inputs
274(1)
14.6 Bottom up saliency
275(1)
14.7 Lateral interactions shape receptive fields
276(1)
14.8 Top-down selective attention vs feedforward processing
277(1)
14.9 Topological maps to simplify connectivity
278(1)
14.10 Biologically decodable output representations
279(1)
14.11 Highlights
279(2)
15 Emotion, motivation, reward value, pleasure, and their mechanisms 281(24)
15.1 Emotion, reward value, and their evolutionary adaptive utility
281(2)
15.2 Motivation and reward value
283(1)
15.3 Principles of cortical design for emotion and motivation
283(1)
15.4 Objects are first represented independently of reward value
284(2)
15.5 Specialized systems for face identity and expression processing in primates
286(1)
15.6 Unimodal processing to the object level before multimodal convergence
287(1)
15.7 A common scale for reward value
287(1)
15.8 Sensory-specific satiety
287(1)
15.9 Economic value is represented in the orbitofrontal cortex
288(1)
15.10 Neuroeconomics vs classical microeconomics
288(1)
15.11 Output systems influenced by orbitofrontal cortex reward value representations
289(2)
15.12 Decision-making about rewards in the anterior orbitofrontal cortex
291(1)
15.13 Probabilistic emotion-related decision-making
292(1)
15.14 Non-reward, error, neurons in the orbitofrontal cortex
292(4)
15.15 Reward reversal learning in the orbitofrontal cortex
296(5)
15.16 Dopamine neurons and emotion
301(1)
15.17 The explicit reasoning system vs the emotional system
301(1)
15.18 Pleasure
302(1)
15.19 Personality relates to differences in sensitivity to rewards and punishers
302(1)
15.20 Highlights
303(2)
16 Noise in the cortex, stability, psychiatric disease, and aging 305(40)
16.1 Stochastic noise, attractor dynamics, and schizophrenia
305(11)
16.1.1 Introduction
305(2)
16.1.2 A dynamical systems hypothesis of the symptoms of schizophrenia
307(1)
16.1.3 The depth of the basins of attraction: mean-field flow analysis
308(1)
16.1.4 Decreased stability produced by reduced NMDA conductances
309(2)
16.1.5 Increased distractibility produced by reduced NMDA conductances
311(1)
16.1.6 Synthesis: network instability and schizophrenia
312(4)
16.2 Stochastic noise, attractor dynamics, and obsessive-compulsive disorder
316(9)
16.2.1 Introduction
316(1)
16.2.2 A hypothesis about obsessive-compulsive disorder
317(2)
16.2.3 Glutamate and increased depth of the basins of attraction
319(3)
16.2.4 Synthesis on obsessive-compulsive disorder
322(3)
16.3 Stochastic noise, attractor dynamics, and depression
325(10)
16.3.1 Introduction
325(3)
16.3.2 A non-reward attractor theory of depression
328(1)
16.3.3 Evidence consistent with the theory
329(2)
16.3.4 Relation to other brain systems implicated in depression
331(1)
16.3.5 Implications for treatments
332(1)
16.3.6 Mania and bipolar disorder
333(2)
16.4 Stochastic noise, attractor dynamics, and aging
335(8)
16.4.1 NMDA receptor hypofunction
335(3)
16.4.2 Dopamine
338(1)
16.4.3 Impaired synaptic modification
338(1)
16.4.4 Cholinergic function and memory
339(4)
16.5 Highlights
343(2)
17 Syntax and Language 345(19)
17.1 Neurodynamical hypotheses about language and syntax
345(6)
17.1.1 Binding by synchrony?
345(1)
17.1.2 Syntax using a place code
346(1)
17.1.3 Temporal trajectories through a state space of attractors
347(1)
17.1.4 Hypotheses about the implementation of language in the cerebral cortex
347(4)
17.2 Tests of the hypotheses - a model
351(4)
17.2.1 Attractor networks with stronger forward than backward connections
351(2)
17.2.2 The operation of a single attractor network module
353(2)
17.2.3 Spike frequency adaptation mechanism
355(1)
17.3 Tests of the hypotheses - findings with the model
355(4)
17.3.1 A production system
355(1)
17.3.2 A decoding system
356(3)
17.4 Evaluation of the hypotheses
359(4)
17.5 Highlights
363(1)
18 Evolutionary trends in cortical design and principles of operation 364(21)
18.1 Introduction
364(1)
18.2 Different types of cerebral neocortex: towards a computational understanding
364(12)
18.2.1 Neocortex or isocortex
365(6)
18.2.2 Olfactory (pyriform) cortex
371(3)
18.2.3 Hippocampal cortex
374(2)
18.3 Addition of areas in the neocortical hierarchy
376(2)
18.4 Evolution of the orbitofrontal cortex
378(1)
18.5 Evolution of the taste and flavour system
379(2)
18.5.1 Principles
379(1)
18.5.2 Taste processing in rodents
380(1)
18.6 Evolution of the temporal lobe cortex
381(1)
18.7 Evolution of the frontal lobe cortex
382(1)
18.8 Highlights
382(3)
19 Genetics and self-organization build the cortex 385(21)
19.1 Introduction
385(1)
19.2 Hypotheses about the genes that build cortical neural networks
386(4)
19.3 Genetic selection of neuronal network parameters
390(1)
19.4 Simulation of the evolution of neural networks using a genetic algorithm
391(10)
19.4.1 The neural networks
391(1)
19.4.2 The specification of the genes
392(5)
19.4.3 The genetic algorithm, and general procedure
397(1)
19.4.4 Pattern association networks
398(2)
19.4.5 Autoassociative networks
400(1)
19.4.6 Competitive networks
400(1)
19.5 Evaluation of the gene-based evolution of single-layer networks
401(2)
19.6 The gene-based evolution of multi-layer cortical systems
403(1)
19.7 Highlights
404(2)
20 Cortex versus basal ganglia design for selection 406(10)
20.1 Systems-level architecture of the basal ganglia
406(2)
20.2 What computations are performed by the basal ganglia?
408(2)
20.3 How do the basal ganglia perform their computations?
410(3)
20.4 Comparison of selection in the basal ganglia and cerebral cortex
413(2)
20.5 Highlights
415(1)
21 Sleep and Dreaming 416(4)
21.1 Is sleep necessary for cortical function?
416(1)
21.2 Is sleep involved in memory consolidation?
417(1)
21.3 Dreams
418(1)
21.4 Highlights
419(1)
22 Which cortical computations underlie consciousness? 420(35)
22.1 Introduction
420(1)
22.2 A Higher-Order Syntactic Thought (HOST) theory of consciousness
421(11)
22.2.1 Multiple routes to action
421(2)
22.2.2 A computational hypothesis of consciousness
423(2)
22.2.3 Adaptive value of processing that is related to consciousness
425(1)
22.2.4 Symbol grounding
426(2)
22.2.5 Qualia
428(1)
22.2.6 Pathways
429(1)
22.2.7 Consciousness and causality
430(1)
22.2.8 Consciousness and higher-order syntactic thoughts
431(1)
22.3 Selection between conscious vs unconscious decision-making systems
432(9)
22.3.1 Dual major routes to action: implicit and explicit
432(7)
22.3.2 The Selfish Gene vs The Selfish Phenotype
439(1)
22.3.3 Decision-making between the implicit and explicit systems
440(1)
22.4 Determinism
441(1)
22.5 Free will
442(1)
22.6 Content and meaning in representations
443(2)
22.7 The causal role of consciousness and the relation between the mind and the brain
445(2)
22.8 Comparison with other theories of consciousness
447(6)
22.8.1 Higher-order thought theories
447(2)
22.8.2 Oscillations and temporal binding
449(1)
22.8.3 A high neural threshold for information to reach consciousness
450(1)
22.8.4 James-Lange theory and Damasio's somatic marker hypothesis
451(1)
22.8.5 LeDoux's approach to emotion and consciousness
451(1)
22.8.6 Panksepp's approach to emotion and consciousness
452(1)
22.8.7 Global workspace theories of consciousness
452(1)
22.8.8 Monitoring and consciousness
452(1)
22.9 Highlights
453(2)
23 Cerebellar cortex 455(8)
23.1 Introduction
455(1)
23.2 Architecture of the cerebellum
456(4)
23.2.1 The connections of the parallel fibres onto the Purkinje cells
456(1)
23.2.2 The climbing fibre input to the Purkinje cell
457(1)
23.2.3 The mossy fibre to granule cell connectivity
457(3)
23.3 Modifiable synapses of parallel fibres onto Purkinje cell dendrites
460(1)
23.4 The cerebellar cortex as a perceptron
460(1)
23.5 Highlights: differences between cerebral and cerebellar cortex microcircuitry
461(2)
24 The hippocampus and memory 463(91)
24.1 Introduction
463(1)
24.2 Systems-level functions of the hippocampus
464(22)
24.2.1 Systems-level anatomy
465(2)
24.2.2 Evidence from the effects of damage to the hippocampus
467(1)
24.2.3 The necessity to recall information from the hippocampus
468(2)
24.2.4 Systems-level neurophysiology of the primate hippocampus
470(8)
24.2.5 Head direction cells in the presubiculum
478(1)
24.2.6 Perirhinal cortex, recognition memory, and long-term familiarity memory
479(7)
24.3 A theory of the operation of hippocampal circuitry as a memory system
486(45)
24.3.1 Hippocampal circuitry
487(1)
24.3.2 Entorhinal cortex
488(2)
24.3.3 CA3 as an autoassociation memory
490(19)
24.3.4 Dentate granule cells
509(6)
24.3.5 CA1 cells
515(1)
24.3.6 Recoding in CA1 to facilitate retrieval to the neocortex
515(5)
24.3.7 Backprojections to the neocortex, memory recall, and consolidation
520(3)
24.3.8 Backprojections to the neocortex - quantitative aspects
523(3)
24.3.9 Simulations of hippocampal operation
526(2)
24.3.10 The learning of spatial view and place cell representations
528(1)
24.3.11 Linking the inferior temporal visual cortex to spatial view and place cells
529(2)
24.3.12 A scientific theory of the art of memory: scientia artis memoriae
531(1)
24.4 Tests of the theory of hippocampal cortex operation
531(15)
24.4.1 Dentate gyrus (DG) subregion of the hippocampus
531(4)
24.4.2 CA3 subregion of the hippocampus
535(7)
24.4.3 CA1 subregion of the hippocampus
542(4)
24.5 Evaluation of the theory of hippocampal cortex operation
546(6)
24.5.1 Tests of the theory by hippocampal system subregion analyses
546(2)
24.5.2 Comparison with other theories of hippocampal function
548(4)
24.6 Highlights
552(2)
25 Invariant visual object recognition learning 554(120)
25.1 Introduction
554(1)
25.2 Invariant representations of faces and objects in the inferior temporal visual cortex
555(21)
25.2.1 Processing to the inferior temporal cortex in the primate visual system
555(1)
25.2.2 Translation invariance and receptive field size
556(1)
25.2.3 Reduced translation invariance in natural scenes
557(3)
25.2.4 Size and spatial frequency invariance
560(1)
25.2.5 Combinations of features in the correct spatial configuration
561(1)
25.2.6 A view-invariant representation
562(3)
25.2.7 Learning in the inferior temporal cortex
565(3)
25.2.8 Distributed encoding
568(4)
25.2.9 Face expression, gesture, and view
572(1)
25.2.10 Specialized regions in the temporal cortical visual areas
572(4)
25.3 Approaches to invariant object recognition
576(6)
25.3.1 Feature spaces
577(1)
25.3.2 Structural descriptions and syntactic pattern recognition
578(2)
25.3.3 Template matching and the alignment approach
580(1)
25.3.4 Invertible networks that can reconstruct their inputs
581(1)
25.3.5 Feature hierarchies
582(1)
25.4 Hypotheses about object recognition mechanisms
582(4)
25.5 Computational issues in feature hierarchies
586(77)
25.5.1 The architecture of VisNet
587(9)
25.5.2 Initial experiments with VisNet
596(7)
25.5.3 The optimal parameters for the temporal trace used in the learning rule
603(1)
25.5.4 Different forms of the trace learning rule, and error correction
604(8)
25.5.5 The issue of feature binding, and a solution
612(12)
25.5.6 Operation in a cluttered environment
624(7)
25.5.7 Learning 3D transforms
631(5)
25.5.8 Capacity of the architecture, and an attractor implementation
636(7)
25.5.9 Vision in natural scenes - effects of background versus attention
643(8)
25.5.10 The representation of multiple objects in a scene
651(2)
25.5.11 Learning invariant representations using spatial continuity
653(1)
25.5.12 Lighting invariance
654(2)
25.5.13 Invariant global motion in the dorsal visual system
656(1)
25.5.14 Deformation-invariant object recognition
656(1)
25.5.15 Learning invariant representations of scenes and places
657(2)
25.5.16 Finding and recognising objects in natural scenes
659(4)
25.6 Further approaches to invariant object recognition
663(6)
25.6.1 Other types of slow learning
663(1)
25.6.2 HMAX
663(5)
25.6.3 Sigma-Pi synapses
668(1)
25.6.4 Deep learning
668(1)
25.7 Visuo-spatial scratchpad memory, and change blindness
669(1)
25.8 Processes involved in object identification
670(1)
25.9 Highlights
671(3)
26 Synthesis 674(20)
26.1 Principles of cortical operation, not a single theory
674(1)
26.2 Levels of explanation, and the mind-brain problem
674(2)
26.3 Brain computation compared to computation on a digital computer
676(5)
26.4 Understanding how the brain works
681(2)
26.5 Synthesis on principles of operation of the cerebral cortex
683(9)
26.5.1 Hierarchical organization
683(1)
26.5.2 Localization of function
684(1)
26.5.3 Recurrent collaterals and attractor networks
684(1)
26.5.4 The noisy cortex
685(1)
26.5.5 Top-down attention
685(1)
26.5.6 Diluted connectivity
685(1)
26.5.7 Sparse distributed graded firing rate encoding
685(1)
26.5.8 Synaptic modification
686(1)
26.5.9 Adaptation and facilitation
686(1)
26.5.10 Backprojections
686(1)
26.5.11 Neurogenesis
687(1)
26.5.12 Binding and syntax
687(1)
26.5.13 Evolution of the cerebral cortex
687(1)
26.5.14 Genetic specification of cortical design
687(1)
26.5.15 The cortical systems for emotion
688(1)
26.5.16 Memory systems
688(1)
26.5.17 Visual cortical processing for invariant visual object recognition
689(1)
26.5.18 Cortical lamination, operation, and evolution
689(3)
26.6 Highlights
692(2)
A Introduction to linear algebra for neural networks 694(12)
A.1 Vectors
694(6)
A.1.1 The inner or dot product of two vectors
694(1)
A.1.2 The length of a vector
695(1)
A.1.3 Normalizing the length of a vector
696(1)
A.1.4 The angle between two vectors: the normalized dot product
696(1)
A.1.5 The outer product of two vectors
697(1)
A.1.6 Linear and non-linear systems
698(1)
A.1.7 Linear combinations, linear independence, and linear separability
699(1)
A.2 Application to understanding simple neural networks
700(6)
A.2.1 Capability and limitations of single-layer networks
701(2)
A.2.2 Non-linear networks: neurons with non-linear activation functions
703(1)
A.2.3 Non-linear networks: neurons with non-linear activations
704(2)
B Neuronal network models 706(109)
B.1 Introduction
706(1)
B.2 Pattern association memory
706(17)
B.2.1 Architecture and operation
707(3)
B.2.2 A simple model
710(2)
B.2.3 The vector interpretation
712(1)
B.2.4 Properties
713(3)
B.2.5 Prototype extraction, extraction of central tendency, and noise reduction
716(1)
B.2.6 Speed
716(1)
B.2.7 Local learning rule
717(5)
B.2.8 Implications of different types of coding for storage in pattern associators
722(1)
B.3 Autoassociation or attractor memory
723(11)
B.3.1 Architecture and operation
724(1)
B.3.2 Introduction to the analysis of the operation of autoassociation networks
725(2)
B.3.3 Properties
727(6)
B.3.4 Use of autoassociation networks in the brain
733(1)
B.4 Competitive networks, including self-organizing maps
734(22)
B.4.1 Function
734(1)
B.4.2 Architecture and algorithm
735(1)
B.4.3 Properties
736(5)
B.4.4 Utility of competitive networks in information processing by the brain
741(2)
B.4.5 Guidance of competitive learning
743(2)
B.4.6 Topographic map formation
745(4)
B.4.7 Invariance learning by competitive networks
749(2)
B.4.8 Radial Basis Function networks
751(1)
8.4.9 Further details of the algorithms used in competitive networks
752(4)
B.5 Continuous attractor networks
756(11)
8.5.1 Introduction
756(2)
B.5.2 The generic model of a continuous attractor network
758(1)
B.5.3 Learning the synaptic strengths in a continuous attractor network
759(2)
B.5.4 The capacity of a continuous attractor network: multiple charts
761(1)
B.5.5 Continuous attractor models: path integration
761(3)
B.5.6 Stabilization of the activity packet within a continuous attractor network
764(2)
B.5.7 Continuous attractor networks in two or more dimensions
766(1)
B.5.8 Mixed continuous and discrete attractor networks
767(1)
B.6 Network dynamics: the integrate-and-fire approach
767(14)
B.6.1 From discrete to continuous time
768(1)
B.6.2 Continuous dynamics with discontinuities
769(4)
B.6.3 An integrate-and-fire implementation
773(1)
B.6.4 The speed of processing of attractor networks
774(3)
B.6.5 The speed of processing of a four-layer hierarchical network
777(3)
8.6.6 Spike response model
780(1)
B.7 Network dynamics: introduction to the mean-field approach
781(2)
B.8 Mean-field based neurodynamics
783(8)
8.8.1 Population activity
783(2)
8.8.2 The mean-field approach used in a model of decision-making
785(2)
8.8.3 The model parameters used in the mean-field analyses of decision-making
787(1)
8.8.4 A basic computational module based on biased competition
788(1)
B.8.5 Multimodular neurodynamical architectures
789(2)
B.9 Sequence memory implemented by adaptation in an attractor network
791(1)
B.10 Error correction networks
792(7)
B.10.1 Architecture and general description
792(1)
B.10.2 Generic algorithm for a one-layer error correction network
793(1)
B.10.3 Capability and limitations of single-layer error-correcting networks
793(4)
B.10.4 Properties
797(2)
B.11 Error backpropagation multilayer networks
799(4)
B.11.1 Introduction
799(1)
B.11.2 Architecture and algorithm
799(3)
8.11.3 Properties of multilayer networks trained by error backpropagation
802(1)
B.12 Biologically plausible networks vs backpropagation
803(1)
B.13 Convolution networks
804(2)
B.14 Contrastive Hebbian learning: the Boltzmann machine
806(1)
B.15 Deep Belief Networks
807(1)
B.16 Reinforcement learning
807(7)
B.16.1 Associative reward-penalty algorithm of Barto and Sutton
808(2)
B.16.2 Reward prediction error or delta rule learning, and classical conditioning
810(1)
B.16.3 Temporal Difference (TD) learning
811(3)
8.17 Highlights
814(1)
C Information theory, and neuronal encoding 815(66)
C.1 Information theory
816(8)
C.1.1 The information conveyed by definite statements
816(1)
C.1.2 Information conveyed by probabilistic statements
817(1)
C.1.3 Information sources, information channels, and information measures
818(1)
C.1.4 The information carried by a neuronal response and its averages
819(3)
C.1.5 The information conveyed by continuous variables
822(2)
C.2 The information carried by neuronal responses
824(14)
C.2.1 The limited sampling problem
824(1)
C.2.2 Correction procedures for limited sampling
825(1)
C.2.3 The information from multiple cells: decoding procedures
826(4)
C.2.4 Information in the correlations between cells: a decoding approach
830(5)
C.2.5 Information in the correlations between cells: second derivative approach
835(3)
C.3 Information theory results
838(41)
C.3.1 The sparseness of the distributed encoding used by the brain
839(11)
C.3.2 The information from single neurons
850(2)
C.3.3 The information from single neurons: temporal codes versus rate codes
852(2)
C.3.4 The information from single neurons: the speed of information transfer
854(12)
C.3.5 The information from multiple cells: independence versus redundancy
866(4)
C.3.6 Should one neuron be as discriminative as the whole organism?
870(1)
C.3.7 The information from multiple cells: the effects of cross-correlations
871(4)
C.3.8 Conclusions on cortical neuronal encoding
875(4)
C.4 Information theory terms - a short glossary
879(1)
C.5 Highlights
880(1)
D Simulation software for neuronal network models 881(9)
D.1 Introduction
881(1)
D.2 Autoassociation or attractor networks
881(3)
D.2.1 Running the simulation
881(2)
D.2.2 Exercises
883(1)
D.3 Pattern association networks
884(2)
D.3.1 Running the simulation
884(2)
D.3.2 Exercises
886(1)
D.4 Competitive networks and Self-Organizing Maps
886(3)
D.4.1 Running the simulation
886(2)
D.4.2 Exercises
888(1)
D.5 Highlights
889(1)
References 890(60)
Index 950
Professor Edmund T. Rolls performs full-time research at the Oxford Centre for Computational Neuroscience, and at the University of Warwick where he is Professor of Computational Neuroscience, and has acted as Professor of Experimental Psychology at the University of Oxford, and as Fellow and Tutor of Corpus Christi College, Oxford. His research links neurophysiological and computational neuroscience approaches to human functional neuroimaging and neuropsychological studies in order to provide a fundamental basis for understanding human brain function and its disorders.