Deep Learning For Sequence Data

Deep learning models can take raw text as input.

• True

• False

............................. is the process of transforming the text into numeric tensors.

• Tokenization

• Vectorization

• Embedding

• Bag-of-words representation

Vectors obtained through ...................... are binary, sparse (mostly made of zeros) and very high dimensional.

• One-hot encoding

• Word-embedding

Vectors created by TF-IDF (Term Frequency-Inverse Document Frequency) can give semantic information.

• TRUE

• FALSE

CNNs use recurrent connections to maintain the memory of past inputs.

• TRUE

• FALSE