• Performed word preprocessing such as noisy text and stopwords removal.
• Conducted feature transformation to convert text data into numerical features using Tokenizer before converting the generated tokens into sequences and applying padding so that each input sequence became in the same length.
• Used Sequential API in Keras including the Embedding and LSTM Layer.
• Utilized GloVe Embeddings to build the embedding matrix for the Embedding Layer.
• Achieved nearly 1.00 in F1 score in classifying real or fake news.

Link to Github Repo

• Deep Learning
• Classification