Search⌘ K
AI Features
Home
Courses
Building Advanced Deep Learning and NLP Projects

Project Creation: Part One

Explore the process of building a recurrent neural network for sentiment analysis on IMDB movie reviews. This lesson guides you through loading the preprocessed dataset, converting numeric data back into text, applying padding to standardize review length, and preparing data for training with embedding layers.

We'll cover the following...

Introduction to the project

Welcome to this project. We will build an RNN based model and perform sentiment analysis on the IMDB movie reviews dataset. In this project, we will be using a simple RNN model and an embedding layer, which we discussed in the previous lessons.

For this project, we will be using the dataset provided by Keras. This dataset is already preprocessed, and the text (reviews) is already converted to its numeric representation.

Let’s start our project by first loading the dataset.

Load the dataset

We will need to have Keras installed for this project. You should have Keras already installed if you made it to this chapter.

Python 3.5
from tensorflow.python.keras.datasets import imdb
((XT,YT),(Xt,Yt)) = imdb.load_data(num_words=10000)
print('Dataset Loaded!')

Explanation:

  • On line 1, we imported our package required for loading the dataset.
  • On line 3, we loaded the dataset by specifying a parameter, num_words = 10,000, which says that we only want the top 10,000 words that most frequently occurred in
...