Deep Learning Basics Lecture2: Neural Networks & Multi-Layer Perceptron

Neural Networks

• "Neural networks are computing system as vaguely inspired by the biological neural networks that constitute animal brains."
• Neural networks are function approximators that stack affine transformations followed by nonlinear transformations.

ex) GoogLeNet, ResNet

 

Linear Neural Networks

• We compute the partial derivatives w.r.t the optimization variables.
• Then, we iteratively update the optimization variables. 편미분을 구하고, 이후에 빼주는 식으로 계산.
• of course, we can handle multi dimensional input and output

one way of interpreting a matrix is to regard it as a mapping between two vector spaces.

 

Multi-Layer Perceptron

• Regression Task
• True target, Predicted output
• Classification Task
• Probabilistic Task