From Associations to Rules: Connectionist Models of Behavior and Cognition: Proceedings of the Tenth Neural Computation and Psychology Workshop (Progress ... Processing) (Progress in Neural Processing)

CONTENTS......Page 8
Introduction......Page 6
Section I High-Level Cognition......Page 12
1. Introduction......Page 14
2. Existing Models of Sequential Control......Page 16
3. The Goal Circuit Model......Page 17
3.1. Goal Units and Simulated Reinforcement Learning......Page 18
3.2. Progressive Routinisation......Page 19
4. Results......Page 20
4.1. The Coffee or Tea Routine......Page 21
4.2. Redundancy and Goal Directed Behaviour......Page 22
4.3. Deliberate Control......Page 23
5. Discussion......Page 24
References......Page 26
1. Introduction......Page 27
1.1. Studies of Contingency Judgment......Page 28
1.2. Illusory Correlation......Page 29
2.1. An Associative Account: The Rescorla-Wagner Model......Page 30
2.2.1. Translating the Problem into “Neural Networks Language”......Page 32
2.2.2. Three-layer Hetero-associative Network......Page 34
2.2.3. Three-layer Auto-hetero-associative Network......Page 35
3. Conclusion......Page 36
References......Page 37
1. Introduction......Page 39
2.1. Theoretical accounts of false memories......Page 40
2.2. DRM findings related to encoding/retrieval......Page 41
3.1. Context models......Page 42
3.2. Contextual retrieval analysis......Page 44
4.2. Results......Page 45
4.3. Interpretation......Page 47
5. Discussion......Page 48
References......Page 49
1. Introduction......Page 52
2. Evolving Neural Network Systems......Page 54
3. Simulation Results for Different Protection Periods......Page 56
4. Allowing the Protection Period to Evolve......Page 57
5. Analysis of the Evolved Performance......Page 59
6. Discussion and Conclusions......Page 61
References......Page 63
Section II Language......Page 64
1. Introduction......Page 66
2. Overview of Child Language at the One-Word and Two-Word Stages......Page 67
3.1. The Modified Counterpropagation Network......Page 68
3.2. One-Word Stage Model Simulation......Page 69
4.1. Temporal Hypermap Model for Two-word Simulation......Page 71
4.2. A Gated Multi-net Architecture for One-Word to Two-Word Transition -......Page 72
4.3. Gated Multinet Model Simulation......Page 73
5. Conclusion and Future work......Page 74
References......Page 75
1. Introduction......Page 77
2. Method......Page 78
2.1. Training the Unimodal Maps......Page 79
2.2.2. On the importance of real acoustic token......Page 80
2.3. Forming the cross-modal associations......Page 81
3.1. Generalisation as a function of number of pairings......Page 82
3.2. Generalisation as a function of pre-pairing experience......Page 84
3.3. Generalisation capacity as a function of the number of Hebbian associations......Page 85
4. Discussion......Page 86
References......Page 87
1. Introduction......Page 89
2.2. Sentences......Page 91
3.1. The dynamical system......Page 92
3.2. Adjusting word vectors......Page 93
3.4. Measuring paradigmaticity......Page 94
4.3. Word representations......Page 95
5. Discussion......Page 96
References......Page 98
1.1. Principles of the Mapping between Orthography and Phonology......Page 100
2. Neglect Dyslexia and the Reading System......Page 102
2.1. Connectionist Models of Reading and Computational Characteristics of the Mapping......Page 103
2.2.2. Environment......Page 106
2.2.3. Training, Testing and Lesioning......Page 107
2.2.4. Results......Page 108
References......Page 109
1. Introduction......Page 111
2. Encoding BSL sign forms......Page 114
3. Encoding sign semantics......Page 117
4. Distance metrics......Page 118
5. Results......Page 119
6. Discussion......Page 120
References......Page 121
Section III Categorization and Visual Perception......Page 124
1. Introduction......Page 126
2. The ST2 Model......Page 128
3. Connecting Modelling and ERPs......Page 129
3.1. The Blaster and the N2pc......Page 132
4.1. Design......Page 133
4.2. Results and Discussion......Page 135
5. Conclusions......Page 136
References......Page 137
1. Introduction......Page 138
1.1. Category Formation in Infants......Page 139
1.2. Previous Models of Infant Categorization......Page 141
2. An Integrated Model of Categorization in Infancy......Page 142
2.1. Data......Page 143
2.2.1. Development of long-term representations......Page 144
2.2.2. The effect of background knowledge on familiarization......Page 146
3. Discussion......Page 147
References......Page 148
1. Introduction......Page 150
2. Background......Page 151
3.1. System Overview......Page 152
3.2. Attention Mechanism......Page 154
4. Databases......Page 156
5. Experiments......Page 157
5.2. Comparing the Performance of Cengji with and without Attention Mechanism......Page 158
6. Conclusion and Future Work......Page 160
References......Page 161
Section IV Sensory and Attentional Processing......Page 162
1. Introduction......Page 164
2. Echo State Networks......Page 165
3.1. Organization of the RNN State Space Initialized with Small Weights......Page 167
3.2. Models Using Markovian Architectural Bias Property......Page 168
3.3. Variable Length Markov Models......Page 170
4.1. Method and Datasets......Page 171
4.2. Reservoir of Echo State Networks......Page 172
4.3. Comparison of ESNs, VLMMS and Models Explicitly Using Markovian Architectural Bias Property......Page 173
References......Page 175
1. Introduction......Page 176
2.1. Head Direction Cells......Page 177
2.2. Sensory Inputs......Page 178
2.3.2. Lateral Mammillary Nucleus (LMN)......Page 179
3. Neural Modelling......Page 180
3.1. Continuous Attractors......Page 181
3.2. Existing Models......Page 182
3.3. The Extended Model......Page 183
3.5.1. Tracking Vestibular Angular Velocity......Page 184
3.5.2. Optic Flow Integration......Page 185
4. Conclusions......Page 186
References......Page 187
1.1.1. Music and syntax......Page 189
1.1.2. The harmonic priming paradigm......Page 190
1.1.3. Sensory and cognitive components in harmonic priming......Page 191
2.1. The present model......Page 192
2.1.1. The auditory peripheral model......Page 193
2.1.3. Sequential learning and the Recursive Self-Organizing Map......Page 195
2.2. The training corpus......Page 197
2.4.1. Topographic representation of periodicity informations......Page 198
3. Conclusions......Page 199
References......Page 200
1. Introduction......Page 202
2. Luminance-chrominance decomposition......Page 203
3. Model of spatio-chromatic sampling by a mosaic of chromatic samples......Page 204
4. Method......Page 207
5. Results......Page 208
6. Discussion......Page 210
References......Page 211
1. Introduction......Page 212
2.1. Path Length, Clustering Coefficient, and the Small-world Network......Page 213
2.2. Global and Local Efficiency of the Network......Page 214
3. The Connectivity of the Real Mammalian Cortex......Page 215
4.1. Dynamics......Page 216
4.3. Performance Measures......Page 217
4.4.1. The W-S Small-world Network......Page 218
5.1. The W-S Small-world Network......Page 219
5.2. The Modular Network......Page 220
6. Conclusion and Discussion......Page 221
References......Page 222
1. Introduction......Page 224
2. Simulation of the primary visual system......Page 225
3.1. Hypothesis......Page 226
3.2. Material and method......Page 227
3.3. Results......Page 228
4.2. Material and method......Page 230
4.3. Results and discussion......Page 231
5. Conclusion......Page 232
6. Acknowledgments......Page 233
7. References......Page 234
AUTHOR INDEX......Page 236