Statistics for Data Science

Analyzing the data and/or applying machine learning to the data

In this phase, quite a bit of analysis takes place as the data scientist (driven by a high level of scientific curiosity and experience) attempts to shape a story based upon an observation or the interpretation of their understanding of the data (up to this point). The data scientist continues to slice and dice the data, using analytics or BI packages—such as Tableau or Pentaho or an open source solution such as R or Python—to create a concrete data storyline. Once again, based on these analysis results, the data scientist may elect to again go back to a prior phase, pulling new data, processing and reprocessing, and creating additional visualizations. At some point, when appropriate ...