Welcome back to the round-up, an overview of the most interesting data science, statistics, and analytics articles of the past week. This week, we have 4 fascinating articles ranging in topics from improving literacy to building brains. In this week's round-up:
- How Unbound Concepts Could Help Improve Literacy
- Machine Learning From Streaming Data
- Data Audits and Data Strategies
- Building Brains to Understand the World’s Data
This is an interesting GigaOM article about how a local startup, Unbound Concepts, has built technology that can understand the contents of childrens books and assigns them a reading level so that kids can be matched more accurately with books appropriate for their reading abilities. The article provides some background about the company, some quotes from their CEO and CTO, and describes some of the features of their products.
This post on the BigML blog tackles the subject of applying machine learning to streaming data. The author distinguishes between time series prediction and non-stationary data distribution problems, provides a couple approaches to solving the latter, and mentions a few considerations you should think about when doing so.
This is an insightful post on Cathy O'Neil's blog about things companies should consider before spending resources on a data science team. Cathy breaks these into two sets of questions she calls a data audit and a data strategy. The questions in the data audit revolve around things like what type of company you're running, what types of data you're collecting, and goals the business has. The data strategy questions have to do more with what you'll actually need to do with the data you've collected and once you've gotten a sense of direction from the audit - more technically oriented questions about models, algorithms, scalability, etc. that will help you decide what kind of team you need to put together.
This is a video of a presentation given by Jeff Hawkins (the founder of Palm, Handspring, and Numenta) about recent advances in understanding how the neocortex works and how Numenta is applying this understanding through sparse distributed representations. Hawkins predicts in the video that the future of intelligent machines is sparse distributed representations and talks about a product called Grok that Numenta has built which uses them to process high velocity machine-generated data.
That's it for this week. Make sure to come back next week when we’ll have some more interesting articles! If there's something we missed, feel free to let us know in the comments below.