Rule #4
Data Science is Cyclic
Abstract
It is a mistake to view a data science problem as just a management exercise or just a technical issue. Some of the infrastructure required for success cannot be mandated or bought—some issues require attitudinal changes within the personnel involved in the project. In this part of the book Data Science for Software Engineering: Sharing Data and Models, some of those changes are listed. They include (1) talk to your users more than your algorithms; (2) know your maths and tools but, just as important, know your problem domain; (3) suspect all collected data and, for all such data, “rinse before use”; (4) data science is a cyclic process, which means you will get it wrong (at first) along ...
Get Sharing Data and Models in Software Engineering now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
