§16§§This two volume set LNCS 4668 and LNCS 4669 constitutes the refereed proceedings of the 17th International Conference on Artificial Neural Networks, ICANN 2007, held in Porto, Portugal, in September 2007.§§The 197 revised full papers presented were carefully reviewed and selected from 376 submissions. The 98 papers of the first volume are organized in topical sections on learning theory, advances in neural network learning methods, ensemble learning, spiking neural networks, advances in neural network architectures neural network technologies, neural dynamics and complex systems, data analysis, estimation, spatial and spatio-temporal learning, evolutionary computing, meta learning, agents learning, complex-valued neural networks, as well as temporal synchronization and nonlinear dynamics in neural networks.§ §04§Learning Theory.- Advances in Neural Network Learning Methods.- Ensemble Learning.- Spiking Neural Networks.- Advances in Neural Network Architectures.- Neural Dynamics and Complex Systems.- Data Analysis.- Estimation.- Spatial and Spatio-Temporal Learning.- Evolutionary Computing.- Meta Learning, Agents Learning.- Complex-Valued Neural Networks (Special Session).- Temporal Synchronization and Nonlinear Dynamics in Neural Networks (Special Session). §04§ Identification.- Advances in Neural Network Learning Methods.- Structure Learning with Nonparametric Decomposable Models.- Recurrent Bayesian Reasoning in Probabilistic Neural Networks.- Resilient Approximation of Kernel Classifiers.- Incremental Learning of Spatio-temporal Patterns with Model Selection.- Accelerating Kernel Perceptron Learning.- Analysis and Comparative Study of Source Separation Performances in Feed-Forward and Feed-Back BSSs Based on Propagation Delays in Convolutive Mixture.- Learning Highly Non-separable Boolean Functions Using Constructive Feedforward Neural Network.- A Fast Semi-linear Backpropagation Learning Algorithm.- Improving the GRLVQ Algorithm by the Cross Entropy Method.- Incremental and Decremental Learning for Linear Support Vector Machines.- An Efficient Method for Pruning the Multilayer Perceptron Based on the Correlation of Errors.- Reinforcement Learning for Cooperative Actions in a Partially Observable Multi-agent System.- Input Selection f §04§or Radial Basis Function Networks by Constrained Optimization.- An Online Backpropagation Algorithm with Validation Error-Based Adaptive Learning Rate.- Adaptive Self-scaling Non-monotone BFGS Training Algorithm for Recurrent Neural Networks.- Some Properties of the Gaussian Kernel for One Class Learning.- Improved SOM Learning Using Simulated Annealing.- The Usage of Golden Section in Calculating the Efficient Solution in Artificial Neural Networks Training by Multi-objective Optimization.- Ensemble Learning.- Designing Modular Artificial Neural Network Through Evolution.- Averaged Conservative Boosting: Introducing a New Method to Build Ensembles of Neural Networks.- Selection of Decision Stumps in Bagging Ensembles.- An Ensemble Dependence Measure.- Boosting Unsupervised Competitive Learning Ensembles.- Using Fuzzy, Neural and Fuzzy-Neural Combination Methods in Ensembles with Different Levels of Diversity.- Spiking Neural Networks.- SpikeStream: A Fast and Flexible Simulator o §04§f Spiking Neural Networks.- Evolutionary Multi-objective Optimization of Spiking Neural Networks.- Building a Bridge Between Spiking and Artificial Neural Networks.- Clustering of Nonlinearly Separable Data Using Spiking Neural Networks.- Implementing Classical Conditioning with Spiking Neurons.- Advances in Neural Network Architectures.- Deformable Radial Basis Functions.- Selection of Basis Functions Guided by the L2 Soft Margin.- Extended Linear Models with Gaussian Prior on the Parameters and Adaptive Expansion Vectors.- Functional Modelling of Large Scattered Data Sets Using Neural Networks.- Stacking MF Networks to Combine the Outputs Provided by RBF Networks.- Neural Network Processing for Multiset Data.- The Introduction of Time-Scales in Reservoir Computing, Applied to Isolated Digits Recognition.- Partially Activated Neural Networks by Controlling Information.- CNN Based Hole Filler Template D
