Democratizing Deep Learning, with keras-pandas

tl;dr: keras-pandas allows users to rapidly build and iterate on deep learning models

Deep Learning is transforming corporate America, and is still an area of active research. While deep learning used to be solely the realm of specialized experts using highly specialized code, the barrier to entry is rapidly falling. It's now possible for traditional data scientists to wring value out of Deep Learning, and Deep Learning experts to have a larger impact by creating code assembly lines (pun intended).

With this in mind, over the past few years I have written keras-pandas, which allows users to rapidly build and iterate on deep learning models.

About the project

Getting data formatted and into keras can be tedious, time consuming, and require domain expertise, whether your a veteran or new to Deep Learning. keras-pandas overcomes these issues by (automatically) providing:

  • Data transformations: A cleaned, transformed and correctly formatted X and y (good for keras, sklearn or any other ML platform)
  • Data piping: A correctly formatted keras input, hidden and output layers to quickly start iterating on

These approaches are built on best in class approaches from practitioners, kaggle grand masters, papers, blog posts, and coffee chats, to simple entry point into the world of deep learning, and a strong foundation for deep learning experts.

Getting started

I'd recommend checking out the Quick start guide to get a feel for the package an interface. If you'd like to dive a bit deeper, you can have a look at the examples, or start building out a model on your own data.

During the beta, I've been fortunate to get feedback from users at companies large and small, ranging users at Google to hedge funds, and from finance to education. If something breaks, or you'd like to request a feature, feel free to reach out.