This article summarizes the important knowledge points related to deep learning. It is presented to you through long graphs and PDFs. You are welcome to download PM.
In 15, Google introduced handwriting input in 82 language and upgraded to 100 language last year. But with the rapid development of machine learning, researchers are constantly refactoring the past methods to bring users a faster and more accurate experience.
We introduce a "curve to save the country" approach, which does not directly change the eigenvalues of the weights, but instead parameterizes the neurons, eliminating the original by generating a linear self-looping path.RNNWeight parameter W, which is the long and short time memory unitLSTM(long short-term memory).
In the previous section (how to understandRNN? (Theory)) The course introduces the basic structure of the cyclic neural network. At the same time, such a loop structure will bring some difficulties to the optimization. This paper mainly introduces two simple ways to alleviate the optimization problem of RNN.
A network with a cyclic structure can be called a recurrent neural network.RNNThis problem can be solved very effectively. From a simple theoretical point of view, it can handle sequences of any length and does not need to fix N in advance, which is more flexible.