Tools and Techniques
This series by Tony Ojeda provides a framework for exploring data with Python. In this part, the goal is to create additional categorical fields that will make data easier to explore and allow it to be viewed from new perspectives.
StreamAlert is a real-time data analysis framework with point-in-time alerting. It's unique in that it’s serverless, scalable to TB’s/hour, has automated infrastructure deployment and it’s secure by default. This is an open source framework from the Engineering & Data Science team at Airbnb.
Part three of a popular series from the Machine Learning at Berkeley organization. This part explores neural networks: what they are, how they work, and how to train one. This series relies a lot on visuals to describe concepts, rather than math or code.
Google recently added a new feature to its Maps for Android application that offers predictions about parking availability. This involves a lot of challenges, including variable parking availability, lack of real-time info, and multi-level parking structures. This post on the Google Research Blog describes their approach.
Fun exploration of applying deep learning to Chess. This is easy to follow and includes a GitHub repo of code.