State-of-the-Art Machine Learning Automation with HDT

In its simplest form, our particular problem consists of analyzing historical data about articles and blog posts, to identify features (also called metrics or variables) that are good predictors of blog popularity when combined together, to build a system that can predict the popularity of an article before it gets published. As we have seen in the previous section, the problem consists of predicting pv, the logarithm of unique page views for an article (over some time period), as a function of keywords found in the title, and whether the article in question is a blog or not. Some nodes have a far larger volatility, for instance when one of the keywords has different meanings, such as the word "training", in "training deep learning" (training set) versus "deep learning training" (courses.) It involves training sets, cross-validation, feature selection, binning, and populating hash tables of key-value pairs (referred to here as the nodes).