Exploratory Analysis

  • Data Summarization
  • Different Databases

    • Operational database
      Transcations OLTP (online transactions processing)
    • Datawarehouse (for business applications) Aggregate queries OLAP (online analytic processing)
  • Visualization

Summary Statistics

Percentiles

Mean

Trimmed mean

Having an estimate on the fraction of outliers $p $ in data

If $p=100 \%$, then the result is Median.

Variance

Absolute average deviation (AAD)

More rboust than variance

Median absolute deviation

Correlation coefficient


In [ ]: