- In this project a
word2vecmodel is incorporated to capture semantic features of words and convert them into high dimentional word vectors (embeddings). - After that Stacked Bi-directional Long Short Term Memory (Stacked Bi-LSTM) model utilizes sequential word vectors transformed from CBOW model to represent contextual features.
Cross Entropyis used to find the difference between predicted and real values and calculate loss for each statement. Based on the loss values,adam optimiseris used for optimising the paramters.
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|