and all others whose names I am guilty of not including here. Finally, I would like to thank my parents and my family for their love and encouragements throughout the years. 65 3.3.5 Summary table of computational levels of study 71 3.4 Potential contributions of research 76 3.4.1 Neuroscientific importance 76 3.4.2 Technological importance 77 11 12 contents 4 experiment, model architecture and implementation 79 4.1 Modelled environment and organism 79 4.2 Task/trial structure 81 4.2.1 The Neutral condition 82 4.2.2 The Gap condition 83 4.2.3 The Overlap condition 84 4.3 Model architecture 86 4.3.1 Model of reflex visual attention 86 4.3.2 Implementation framework 86 4.3.3 Neural unit architecture 87 4.3.4 Layer structure and connectivity 94 4.3.5 Biologically plausible learning 103 4.3.6 Learning and testing implementation 107 5 results 111 5.1 Unit behaviour 111 5.2 Network learning 115 5.3 Network behaviour 117 5.3.1 Statistical analysis 117 5.3.2 Network performance 130 6 discussion 133 6.1 Recapitulation of the study 133 6.2 Theories embodied in the model 134 6.3 Results interpretation and model predictions 135 6.4 Research contributions 138 6.5 Limitations of the model 139 6.6 Future experimental research 141 6.6.1 Experimental protocol 141 6.6.2 Related prior study 143 6.7 Future theoretical research 144 6.7.1 Methodological work 145 iii pycogmo: a learning framework for computational neuroscience 147 7 motivation and background 149 7.1 Review of related literature and software 149 7.2 Overview of PyCogMo 152 7.3 Potential contributions of research 153 7.3.1 Synaptic plasticity 153 7.3.2 The problem of synaptic consolidation in PyNN 154 7.3.3 Scientific goals of PyCogMo 155 7.3.4 Design goals of PyCogMo 156 contents 13 8 implementation and testing 159 8.1 System Design 159 8.2 Components design 160 8.2.1 The learning module 160 8.2.2 The scheduling module 163 8.2.3 The visualisation module 165 8.2.4 The utilities component 168 8.3 Implementation and testing 169 8.3.1 Implementation of the scheduling module 169 8.3.2 Implementation of the learning module 172 8.3.3 Implementation of the visualisation module 174 8.3.4 Implementation of the utilities module 177 8.3.5 Testing 178 9 discussion 179 9.1 Issues and limitations 179 9.1.1 System design 179 9.2 Research contributions 181 9.3 Future work 181 iv appendices 185 a principles of astructural neuron modelling 187 a.1 From equivalent circuit models to the Hodgkin-Huxley model 187 a.1.1 Simple electrical circuits 188 a.1.2 Modelling a patch of membrane 188 a.2 From Hudgkin-Huxley to simplified models 196 a.2.1 The Integrate-and-Fire model 197 a.2.2 A more elegant model: The exponential Integrate-and-Fire 198 a.2.3 An most elegant model: the theta neuron 198 a.2.4 An extra dimension 198 a.3 More realistic models 199 b neural simulation software 201 conclusion of the thesis 205 bibliography 207