KDnuggets Home » News :: 2013 :: Nov :: Publications :: Data Science not just for Big Data: Video and Slides ( 13:n28 )

Data Science not just for Big Data: Video and Slides

          

While big data certainly brings changes to data science, major data science principles remain unchanged regardless of the data size. Watch leading experts David Smith from Revolution Analytics and Gregory Piatetsky from KDnuggets discuss key data science principles.

By Gregory Piatetsky, Nov 19, 2013.

Data Science vs Big DataLast month I joined David Smith from Revolution Analytics for an interesting webinar, hosted by Kalido CTO Darren Peirce, on Data Science not just for Big Data (Oct 16, 2013).

While big data certainly brings changes to data science, some key data science principles remain unchanged regardless of the data size.

Among these principles are:

  • Focus on actionable patterns
  • Build predictive models - supervised learning (train, test, x-validate)
  • Avoid overfitting
  • Avoid information leakers
  • Calculating similarity of objects for data-driven clustering, unsupervised learning
  • Select important variables/features
  • Model accuracy vs lift: how much more prevalent a pattern is than would be expected by chance
  • Estimate probability and cost/gain of actions
  • Help optimize decisions

Here are my slides from the webinar

and check the amazing story of Ignaz Semmelweis, one of the first "data scientists".

Here are David Smith slides: Data Science - not just for Big Data from the webinar.

Watch the webinar below.



KDnuggets Home » News :: 2013 :: Nov :: Publications :: Data Science not just for Big Data: Video and Slides ( 13:n28 )