A Data Science Central Community
We all love features... lots of features! ...in our new cars, in our gadgets, in our smart phones, in our toys, and in our data sets!
Consider this toy that we found at the thrift store for under $6.00:
This toy house delivers numerous musical and other sound effects that are triggered whenever one of the features in the house is pressed, or moved, or rotated: the phone, the dog, the bird, the doorbell, the refrigerator, the stove, the sunrise over the roof, the squirrel climbing the tree, and even the bathroom plumbing. It is a playland of fun, distraction, discovery, and amusement for young children (and occasional older children).
We love our data the same way -- full of interesting features to explore and to learn from: patterns, correlations, trends, associations, novel items, new knowledge, etc. Consequently, the era of Big Data, which generates oceans of feature-rich data with big Variety (not only big Volume), can be an exciting time for all of us as we explore those features for new discoveries and business insights.
Therefore, it is no surprise that feature mining is one aspect of data science that appeals to all data scientists. Feature mining includes feature generation (from combinations of existing attributes), feature selection (for mining and knowledge discovery), and feature extraction (for operational systems, decision support, and reuse in various analytics processes, dashboards, and pipelines).
To learn more, check out the recent data science article on "Interactive Visualization enabled Feature Selection and Model Creation".