I am bit confused with the brainjs implementation. If I had to use brain.recurrent.RNN, how can that call be passed whilst adding documents?
If my understanding is correct, this classifier is using node natural only for stemming, stop-words removal right? This is not using the Naive Bayes classifier from the natural library.
From what I saw inside the library, it seems to be a feed-forward network with the default options. I guess if you wanted to provide a different architecture, you'd have to modify this library to allow it.
Hey Man
I am bit confused with the brainjs implementation. If I had to use
brain.recurrent.RNN
, how can that call be passed whilst adding documents?If my understanding is correct, this classifier is using node natural only for stemming, stop-words removal right? This is not using the Naive Bayes classifier from the
natural
library.