Skip to content

Case study: Analyze the candy power ranking to identify and recommend popular candy characteristics

License

Notifications You must be signed in to change notification settings

szapp/CandyAnalysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Candy Analysis

Case study: Analyze the candy power ranking to identify and recommend popular candy characteristics.

The dataset by FiveThirtyEight is distributed un the Creative Commons Attribution 4.0 International license (https://creativecommons.org/licenses/by/4.0/).

Project

The production of a new candy is planned. Among the project team there is no consensus about the characteristics of the candy.

Based on a dataset from market analysis, the task is to give a clear recommendation for what characteristics the new product should express.

Deliverable

The results are compiled in a presentation with a clear recommendation. The presentation is in German, but the numbers speak for themselves.


Libraries used

  • Scipy
  • Scikit-learn
  • Seaborn

See requirements.txt.

Challenges

  • Small dataset (86 rows/samples)
  • Data is aggregated over brands (e.g. win percentage)
  • Study design might not be fair (not blind)

Approach

Treating the problem not as a regression but as a classification and using statistical analysis allows to identify features that are statistically dependent with popular brands. With interaction terms, the combination of successful characteristics can be recommended.

About

Case study: Analyze the candy power ranking to identify and recommend popular candy characteristics

Topics

Resources

License

Stars

Watchers

Forks