I will organise the repository based on the following directory tree:

• Main
  • Methods
    • Autoencoders
    • VariationalAutoencoders
    • NaturalLanguageProcessing
  • Results
    • Plots
    • Summary
  • LiteratureReview

Here, I will add the additional code necessary for moving to the next stage of the project: Data generation.

Here, I am going to the Nietzsche database to implement autoencoders and VAEs. It is a toy example for training for generative modelling applied to text data.

Here, I will start generating new data based on the trained model.

Here, I will start preparing the main framework to use autoencoders with 5G files.

Input:

• Load the CSV file from 5G data;
• Transform the categorical data;
  Output:
• Create the encoders variables;
• Generate More data.

Here, I will start implementing the autoencoders and variational autoencoders, considering the 5G dataset. For that, I will use the data available at:

5Gdataset

Here, I am going to study more methods for implementing autoencoder with text data.

Study sources:

• DataCamp courses
• Research Papers
• Natural Language Processing In action

Here, I am going to use Keras and Tensorflow for the implementation. The dataset MNIST will be used for training the autoencoder and the variational autoencoder (VAE) methods. So, this issue will help the developer to understand more the processes and its limitation during training. Additionally, it will be the initial block for the project.