Difference between revisions of "Exploratory Data Analysis (EDA)"

From InfoVis:Wiki
Jump to navigation Jump to search
Line 2: Line 2:
  
 
{{FloatingQuote|EDA is an approach to data analysis that postpones the usual assumptions about what kind of model the data follow with the more direct approach of allowing the data itself to reveal its underlying structure and model.| [Filliben, 2005]}}
 
{{FloatingQuote|EDA is an approach to data analysis that postpones the usual assumptions about what kind of model the data follow with the more direct approach of allowing the data itself to reveal its underlying structure and model.| [Filliben, 2005]}}
 +
 +
{{Quotation|[...] is concerned primarily with explorations and description of data, not with inference. The techniques are designed to identify fundamental, conceptually
 +
meaningful patterns and relationships in data and to call attention to
 +
observations that deviate greatly from those fundamental patterns| [Smith and Prentice, 1993]}}
  
 
Furthermore, EDA can be used to support the selection of appropriate statistical tools as well as to provide a basis for statistical inference and further data collection.  
 
Furthermore, EDA can be used to support the selection of appropriate statistical tools as well as to provide a basis for statistical inference and further data collection.  
Line 10: Line 14:
  
 
*[Filliben, 2005]: James J. Filliben, NIST/SEMATECH [http://www.itl.nist.gov/div898/handbook/|''e-Handbook of Statistical Methods''], 2005.
 
*[Filliben, 2005]: James J. Filliben, NIST/SEMATECH [http://www.itl.nist.gov/div898/handbook/|''e-Handbook of Statistical Methods''], 2005.
 +
*[Smith and Prentice, 1993]: A.F.Smith and D.A. Prentice, ''Exploratory data analysis'', A handbook for data analysis in the behavioral sciences: Statistical issues, pages 349-390, 1993.
  
 
[[Category: Glossary]]
 
[[Category: Glossary]]

Revision as of 04:01, 23 June 2005

Exploratory data analysis (EDA) was introduced by John Tukey as an approach to analyze data when there is only a low level of knowledge about its cause system as well as contextual information. EDA aims at letting the data itself influence the process of suggesting hypotheses instead of only using it to evaluate given (a priori) hypotheses.
EDA is an approach to data analysis that postpones the usual assumptions about what kind of model the data follow with the more direct approach of allowing the data itself to reveal its underlying structure and model.
[Filliben, 2005]


[...] is concerned primarily with explorations and description of data, not with inference. The techniques are designed to identify fundamental, conceptually

meaningful patterns and relationships in data and to call attention to

observations that deviate greatly from those fundamental patterns
[Smith and Prentice, 1993]


Furthermore, EDA can be used to support the selection of appropriate statistical tools as well as to provide a basis for statistical inference and further data collection.

Essential to EDA are graphical tools like box plots, stem–and–leaf plots, scatter plots, or timelines.

References

  • [Filliben, 2005]: James J. Filliben, NIST/SEMATECH e-Handbook of Statistical Methods, 2005.
  • [Smith and Prentice, 1993]: A.F.Smith and D.A. Prentice, Exploratory data analysis, A handbook for data analysis in the behavioral sciences: Statistical issues, pages 349-390, 1993.