/Statistics

Statistics methods for data science

Primary LanguageJupyter Notebook

Statistics

Statistics methods for data science

1_Confident_Intervals

  • Confidence intervals for the mean
  • Confidence intervals for proportion

2_Hypothesis_Test

  • Parametric and non-parametric criterias

3_Models_Comparison

  • t-statistics for two samples (predictive models) comparison

4_Correlation

  • Correlation analysis for real, binominal, categorical data

5_Multiple_Hypothesis_Test

  • Four models comparison using correction for multiple validation: Holm method with Benjamini–Hochberg method
  • python notebook
  • pandas
  • numpy
  • scipy
  • sklearn
  • statsmodels
  • matplotlib