|
Course Outline
● Psychology and its relation to neural networks ● Models of Artificial Neural Networks ● Learning rules ● Single -layer perceptron classifiers ● Perceptron Convergence Theorem ● Multilayer Feed-forward networks ● Error Back-Propagation Training, and how to enhance it ● Structure selection ● Functional link networks ● Performance Optimization, and Conjugate Gradient Algorithm ● Universal Approximation Theorems, and NNs as Universal Approximators ● The Radial Basis Function Networks ● Single-layer feedback networks ● Hopfield networks (Discrete-Time, and Gradient-Type) ● Recurrent auto-associative memory ● Discrete, Continuous, and Adaptive Bidirectional Associative Memory (BAM) ● Two Coding Strategies for Bidirectional (Multiple training, and Dummy Augmentation) ● Multidirectional Associative Memory (MAM), and Temporal Associative Memory (TAM) ● Neural Networks based on Principal Components Analysis (PCA) ● Hamming nets and MAXNET ● Unsupervised learning of clusters, and Kohonen Network ● Self-Organizing Maps (SOM) ● Learning Vector Quantization (LVQ) ● Counter-Propagation Network (CPN) ● Adaptive Resonance Theory (ART) ● MATLAB Neural Network Toolbox ● Applications of neural algorithms and systems (Control, Robotics, Pattern/Character recognition, Adaptive noise cancellation, etc.) ● New (applied/theoretical) emerging topics will be covered in the final projects
|