CS231n笔记 Lecture 10, Recurrent Neural Networks

Recaption on CNN Architecture

Although Serena is very beautiful, Justin is a better lecturer. Love him.

Recurrent Neural Network

Meant to process sequencial data, reuse hidden state to retain the knowledge of previous fed inputs. Can be use in "one to many", "many to one" and "many to many" scenarios by using different input and output stradegies. Formally, we maintain an $h_t$ for tth iteration, and generate next hidden state by applying $h_{t+1}=f_{W}(h_t, x_{t+1})$ where we reuse the parameters $W$ and get updated through flowing gradients, and get our output from hidden state. A vanilla RNN looks like this: $h_t=\tanh(W_{hh}h_{t-1} + W_{xh}x_t), y_t = W_{hy}h_t$. During training, we backprop the loss through $y_{t}$ and update $W$, for long sequence, this will lead to heavy memory usage (for maintaining the hidden state, as I have experienced), so we chunk the input and update after each chunk.

A very insightful and concrete discussion comes from Andrej Karpathy, The Unreasonable Effectiveness of Recurrent Neural Networks And he provided a sheet of  practical RNN code on gist (a non-exsitent site in China :-( ), it consists only 112 lines of Python.

CS231n笔记 Lecture 10, Recurrent Neural Networks_第1张图片

   CS231n笔记 Lecture 10, Recurrent Neural Networks_第2张图片

Optimization: allows for multiple layers, uses an LSTM instead of a vanilla RNN, has more supporting code for model checkpointing, and is of course much more efficient since it uses mini-batches and can run on a GPU. 

Image Captioning

Andrej's dissertation paper  covers the interesting part of combining image and language, check it out!

The structure here is to remove the last output layers of a convnet (say, the FC-1000 and softmax of a VGG16), and use the result as the hidden state of a RNN by a $W_{ih}$, start with token, keep sampling the out put $y_t$ and pass the previous output to the next input, until we get a token.

CS231n笔记 Lecture 10, Recurrent Neural Networks_第3张图片

 

Image Captioning with Attention

This is a slightly advanced model. First, the CNN produces features instead of sumarization vector (as previously do), and the RNN are allowed to steer its attention on the features (makes them as weighted input), Second, the RNN produces 2 output at each time step, one for attention and one for words.

CS231n笔记 Lecture 10, Recurrent Neural Networks_第4张图片

 

Visual Question Answering

CS231n笔记 Lecture 10, Recurrent Neural Networks_第5张图片

 

RNN allows you to solve some very fancy problems, it is cool!

Multilayer RNN, LSTM and GRU

Deeper models are better, so, stacking layers spacially to get multilayer RNN.

Vanilla RNN Gradient Flow. Leads to Exploding gradients (Solved by Gradient clipping, if L2 norm of gradient to large, .e.g bigger than some threshold, then clip it to that threshold) or Vanishing gradients (Solved by fancier architecture).

 LSTM [Hochreiter and Schmidhuber, 1997]

CS231n笔记 Lecture 10, Recurrent Neural Networks_第6张图片

op in the lecture means element wise multiplication. Note the different non-linearity used and the size of matrix above.

Gradient flow. Uninterrupted, because of c.

玄学结构。。

GRU [Learning phrase representations using rnn encoder-decoder for statistical machine translation, Cho et al. 2014]

[LSTM: A Search Space Odyssey, Greff et al., 2015]

[An Empirical Exploration of Recurrent Network Architectures, Jozefowicz et al., 2015]

Key Idea:  Manage gradient flow.

 

转载于:https://www.cnblogs.com/ichn/p/8503858.html

你可能感兴趣的:(CS231n笔记 Lecture 10, Recurrent Neural Networks)