How to make predictions via LSTM type recurrent neural networks algorithm ???

[chegmarco1989]

Hello.

How to make predictions via LSTM type recurrent neural networks algorithm ???

We don't see anywhere in the documentation how to use LSTM-like recurrent neural networks.

How do LSTM recurrent neural networks work in PHP with this library ???

Please reply to us.

[yuichiis]

Thank you for asking.

See below for general RNN usage

https://rindow.github.io/neuralnetworks/tutorials/neural-machine-translation-with-attention.html

See below for limited use of RNNs in sequential models

https://rindow.github.io/neuralnetworks/tutorials/learn-add-numbers-with-rnn.html

Both usages are the same as tensorflow.

[chegmarco1989]

Thank you for your reply. But that doesn't answer my question.

And according to my research, it is the Recurrent Neural Networks of the LSTM type that make it possible to do this.

So, which of the Recurrent Neural Networks is of LSTM type on Rindow PHP Library and which can allow me to process a set of data and provide precise answers when asked questions ???

[yuichiis]

This is the API list in the library.

https://rindow.github.io/neuralnetworks/api/apitoc.html

I haven't given it a strange name, so as you can see, the following three are RNN layers.

SimpleRNN
LSTM
GRU

Please use it according to the application you are trying to make.

I'm sorry I don't have the ability to answer "basic knowledge of machine learning" or "commentary on artificial intelligence research".

[chegmarco1989]

All right. Thanks.