In the next 2 or 3 years, everybody doing data science should be using a data science productivity platform…a cloud-based data science platform.
In addition to the productivity platforms, the power methods will see some improvements. Here are 2 that Tristan mentions:
- Probabilistic Programming – matching of computer science and Bayesian statistics
- Deep Reinforcement Learning – making optimal decisions via deep learning
It is an exciting time for data science. I think the next few years will see much better productivity tools, workflows, and platforms. More on that in an upcoming blog post.