Feature Quality Estimation

This page depicts the feature quality of features generated with FeGeLOD, using the zoo data set from UCI, the AAUP faculty salary data set (using two target variables), the Mercer quality of living data set, and the Transparency International CPI data set. The figures below show the information gain, the gain ratio, and the chi squared statistic of each feature type.

Zoo Data Set

Information Gain:

Zoo InfoGain

 

Gain Ratio:

Zoo GainRatio

 

Chi Squared Statistic:

Zoo ChiSquared

 AAUP Data Set - Target Variable: Salary of Full Professors

Information Gain:

Colleges Salary InfoGain

Gain Ratio:

Colleges Salary GainRatio

Chi Squared Statistic:

Colleges Salary ChiSquared

AAUP Data Set - Target Variable: Compensation of Full Professors

Information Gain:

Colleges Compensation InfoGain

 

Gain Ratio:

Colleges Compensation GainRatio

 

Chi Squared Statistic:

Colleges Compensation ChiSquared

 

Mercer Quality of Living Data Set

Information Gain:

Mercer InfoGain

 

Gain Ratio:

Mercer GainRatio

 

Chi Squared Statistic:

Mercer ChiSquared

 

Transparency International Data Set

Information Gain:

Transparency InfoGain

 

Gain Ratio:

Transparency GainRatio

 

Chi Squared Statistic:

Transparency ChiSquared

 

 

 
A A A | Drucken | Impressum | Sitemap | Suche | Mobile Version
zum Seitenanfangzum Seitenanfang