1. Model classification
Deep learning is a series of models and methods for solving problems, but deep learning models are not the only research direction in the field of deep learning, and they are not necessarily the most important research direction. In addition to the model, the more important ones are optimization algorithms, loss functions, sampling methods, etc.
1.1 DNN
Deep Neural Networks (hereinafter referred to as DNN) is the basis of deep learning. DNN is sometimes also called multi-layer perceptron (MLP). The neural network layers
inside DNN can be divided into three categories: input layer , hidden layer and output layer, including forward propagation algorithm, back propagation algorithm BP (back propagation)
BP neural network is a multi-layer feedforward neural network trained according to the error back propagation algorithm. Vanilla Neural Networks is actually BP neural network network. Vanilla originally means vanilla, which is basically equivalent to raw, the most original CNN.
1.2 CNN Convolutional neural network
can effectively reduce the dimensionality of large-volume images into small-volume images (without affecting the results); it can retain the characteristics of images, similar to human vision principles; practical applications: image classification, retrieval; target
positioning Detection; target segmentation; face recognition; skeleton recognition.
1.3 RNN Recurrent Neural Network
The structure of traditional neural network: input layer – hidden layer – output layer. The biggest difference between RNN and traditional neural networks is that each time the previous output result is brought to the next hidden layer and trained together. RNN has short-term memory problems and cannot handle very long input sequences; training