신경망 입문

신경과학과 인지과학에서의 대규모 평행 분산 처리 모델을 소개함. 선형 모델, 통계적 패턴 이론, 헤비안 규칙, 자기 조직화, 비선형 모델, 정보 최적화, 신경 정보의 표현 방법 등을 내용으로 함. 또한 이들 모델을 감각 정보 처리, 인지, 학습, 기억 등에 응용함.

Introduction of the course

1.		Introduction of the course
		The Neurons
2.		Neural Models
		Generic neuron model
3.		Lateral inhibition
		Matrices
4.		Linear systems, learning & Memory
		Linear Associator
5.		Non-linear networks, Perception
		Regression, Widrow-Hoff, backprop
6.		Hopfield networks
		Probability and neural networks
7.		Belief Propagation: regression revisited Application to perceptual surface completion
		Utility & probabilty: Bayes decision theory
8.		EM
		Example of large-scale neural architecture. Overview of visual system
9.		Neural coding & representation of visual information (pdf file--second half of last lecture presentation)
		Efficient coding, Self-organization, Principal Components Analysis
10.		Feedforward models in vision: edges
		Feedforward models in vision: material, texture & lightness
11.		Feedforward models in vision: motion
		Integration of visual information
12.		Short presentations of final project plans
		Visual Recognition
13.		Recognition, top-down, feedback models
		Recognition, top-down, feedback model continued