OSEMN framework is Acquire, Scrub, Discover, Mannequin and iNterpret.
Acquire is to assemble information from related sources.
- Decide what information could be helpful
- Consider what information can be found
- Determine on how the information could be gathered
Scrub is to scrub the information to make sure a constant and helpful format.
- Appropriate inconsistent formatting
- Take away duplicate data
- Deal with lacking values
- Take away inaccurate data
Discover is to seek for patterns.
- Study variable distributions
- Study variable relationships
- Carry out statistical exams
Mannequin is to generate predictions and insights.
- Choose a mannequin kind in your targets (typically in cooperation with a companion)
- Classes of fashions embrace:
- Classification — Is that this “A” or “B”?
- Regression — How a lot or what number of?
- Clustering — What pure segments can we discover in our information?
iNterpret is to current and talk your insights.
- Construct visualizations
- Assemble tales
- Create shows of your findings
