Data scientists – the analytical professionals who straddle the line between statistician and software engineer – are in demand like never before. Due to the scarcity of data science talent, it has become increasingly important for data scientists to spend less time answering one-off questions and more time building analytical applications that enable a broad class of users to interact with large data sets, ask detailed questions, and make valid inferences.
This talk gives an overview of the current best practices around creating analytical applications on Hadoop, including dashboards, data APIs, and machine learning models, and then describe how the next generation job scheduling system will enable data scientists to build tools that take Hadoop beyond MapReduce.
Video producer: http://www.meetup.com/Data-Mining/