north0n-FI / Analysing-IMDB-reviews-using-GloVe-and-LSTM

Using the IMDB data found in Keras here a few algorithms built with Keras. The source code is from Francois Chollet's book Deep learning with Python. The aim is to predict whether a review is positive or negative just by analyzing the text. Both self-created as well as pre-trained (GloVe) word embeddings are used. Finally there's a LSTM model and the accuracies of the different algorithms are compared. For the LSTM model I had to cut the data sets of 25.000 sequences by 80% to 5.000, since my laptop's CPU was not able to run the data crunching, making the model's not fully comparable.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Analysing-IMDB-reviews-using-GloVe-and-LSTM

Using the IMDB data found in Keras here a few algorithms built with Keras. The source code is from Francois Chollet's book Deep learning with Python. The aim is to predict whether a review is positive or negative just by analyzing the text. Both self-created as well as pre-trained (GloVe) word embeddings are used. Finally there's a LSTM model and the accuracies of the different algorithms are compared. For the LSTM model I had to cut the data sets of 25.000 sequences by 80% to 5.000, since my laptop's CPU was not able to run the data crunching, making the model's not fully comparable.

About

Using the IMDB data found in Keras here a few algorithms built with Keras. The source code is from Francois Chollet's book Deep learning with Python. The aim is to predict whether a review is positive or negative just by analyzing the text. Both self-created as well as pre-trained (GloVe) word embeddings are used. Finally there's a LSTM model and the accuracies of the different algorithms are compared. For the LSTM model I had to cut the data sets of 25.000 sequences by 80% to 5.000, since my laptop's CPU was not able to run the data crunching, making the model's not fully comparable.


Languages

Language:Jupyter Notebook 100.0%