Chapter 6Architecture and Equilibra结构和平衡
Architecture and Equilibra 结构和平衡 Chapter 6
Chapter 6Architecture and EquilibriaPerfacelyaoynov stabletheoremI：整个系统集合S稳定系统集合工：可由李亚普诺夫函数判定稳定的系统集合2002.12.4
2002.12.4 Chapter 6 Architecture and Equilibria Perface lyaoynov stable theorem I I S L I : 整个系统集合 S : 稳定系统集合 L : 可由李亚普诺夫函数判定稳定的系统集合
Chapter 6Architecture and Eguilibria6.l Neutral Network As Stochastic GradientsystemClassify Neutral network model By their synaptic connection topolgies andby how learning modifies their connectiontopologiessynapticconnectiontopolgies1.feedforward.if No closed synaptic loops2.feedback if closed synapticloopsorfeedback pathwayshowlearning modifiestheirconnectiontopologies1.Supervisedlearning:useclass-membershipinformationoftrainingsamplings2.Unsupervised learning:useunlabelledtrainingsamplings2002.12.4
2002.12.4 Chapter 6 Architecture and Equilibria 6.1 Neutral Network As Stochastic Gradient system Classify Neutral network model By their synaptic connection topolgies and by how learning modifies their connection topologies feedback i f closed synapticloops orfeedback pathways feedforward i f No closed synaptic loops 2. . 1. . − U n ervised learning use unlabelledtrainingsamplings trainingsamplings Supervisedlearning use class membership ormationo f 2. sup : 1. : inf synaptic connection topolgies how learning modifies their connection topologies
Chapter6Architectureand Equilibria6.l Neutral Network As Stochastic GradientsystemDecodeFeedforwardFeedbackOcSGradiedescentLMSRecurrentBackPropagationBackPropagationReinforcement LearingRABAMansudVetor QuantizationBroenian annealingABAMSelf-Organization MapsART-2Competitve learningBAM-Cohen-GrossbergModelCounter-propagationHopfield circuitBrain-state-In BoxAdaptive-ResonanceART-1ART-2Neural NetWork Taxonomy2002.12.4
2002.12.4 Chapter 6 Architecture and Equilibria 6.1 Neutral Network As Stochastic Gradient system Gradiedescent LMS BackPropagation Reinforcement Learing Recurrent BackPropagation Vetor Quantization Self-Organization Maps Competitve learning Counter-propagation RABAM Broenian annealing ABAM ART-2 BAM-Cohen-Grossberg Model Hopfield circuit Brain-state-In_Box Adaptive-Resonance ART-1 ART-2 Feedforward Feedback Decode d e s i v r e p u S d e s i v r e p u s n U e d o c n E Neural NetWork Taxonomy
Chapter 6Architecture and Equilibria6.2Global Equilibra:convergence andstabilityNeural network :synapses,neuronsthree dynamical systems:synapses dynamical systemsMneuons dynamical systemsxjoint synapses-neurons dynamical systems (x,M)Historically,Neural engineers study the first or second neuralnetwork.They usually study learning in feedforward neuralnetworks and neural stability in nonadaptivefeedbackneural networks. RABAM and ART network depend on jointegailibfation of the synaptic and neuronal dynamical systems
2002.12.4 Chapter 6 Architecture and Equilibria 6.2 Global Equilibra:convergence and stability Neural network :synapses , neurons three dynamical systems: synapses dynamical systems neuons dynamical systems joint synapses-neurons dynamical systems Historically,Neural engineers study the first or second neural network.They usually study learning in feedforward neural networks and neural stability in nonadaptive feedback neural networks. RABAM and ART network depend on joint equilibration of the synaptic and neuronal dynamical systems. ' M ' X ( , ) ' ' X M