I'm interested in starting a project which involves using machine learning to purely generate quotes, and then having a Mail Server which sends an email to contacts containing the quote created. I'm trying to expand on the idea of a quote of the day server using machine learning, as I'm new to machine learning and networking and would like to play around with both.
From what I have gathered, I am going to want to use a GAN (generative adversarial network) for generating quotes, and to train a GAN I can follow certain steps:
- Sample a noise set and a real-dataset of the same size
- Train the discriminator on that data
- Sample a different noise subset with the same size
- Train the generator on this data
- Repeat the steps
However, I cannot find much information on how to make this algorithm communicate with a server like the quote of the day mail server I mentioned I wanted to build. I'm wondering if anyone can assist me with figuring out the steps and process involved with the connecting of machine learning to a server like this!
I'm trying to build an outline of what I need to do, any guidance or references would be awesome!