1. 12 September - 18 September
課程內容簡介、學期評分方式說明。
Supervised Learning Network (I)
2.1 Single-Layer Networks (Adalines)
a. Single-Layer Perceptrons (SLPs)
b. Optimization Method (Least-Square Learning Rule )
Supervised Learning Network (II)
2.2 Multi-Layer Networks (Madalines)
a. Multi-Layer Perceptrons (MLPs)
b. Optimization Method ( Back propagation, Conjugate Gradient method, Levenberg-Marquardt (LM) method
2.3 Radial-Basis Networks
Supervised Learning Network (III)
2.4 Cascade-Correlation Networks
2.5 Polynomial Networks
Supervised Learning Network (IV)
2.6 Recurrent Networks (Time series, Back propagation through time, Finite Impulse Response (FIR) MLP ), Temporal Differences method (TD).
Unsupervised Learning Network (I)
3.1 Simple Competitive Networks: Winner-take-all
3.2 Counter propagation Networks (CPN)
Unsupervised Learning Network (II)
3.3 Hamming Network
3.4 Principal Component Analysis and Hebbian Learning (PCA).
Unsupervised Learning Network (III)
3.5 Learning Vector Quantization (LVQ)
3.6 Adaptive Resonance Theory (ART)
Unsupervised Learning Network (III)
3.7 Learning Vector Quantization (LVQ)
3.8 Kohonen Self-Organizing Maps (SOMs)
Others
4.3 Adaptive Neuro-Fuzzy Inference Systems (ANFIS)
4.4 Neural Networks and the Soft Computing