Skip to main content

Table 6 Pearson correlation coefficients (PCC) and Shapiro–Wilk tests

From: Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone

Pearson correlation coefficientShapiro–Wilk test
RankFeatureabs(PCC)RankFeaturep-value
1Serum creatinine0.2941Creatinine phosphokinase7.05×10−28
2Ejection fraction0.2692Serum creatinine5.39×10−27
3Age0.2543Smoking4.58×10−26
4Serum sodium0.1954Death event4.58×10−26
5High blood pressure0.0795Sex1.17×10−25
6Anaemia0.0666High blood pressure1.17×10−25
7Creatinine phosphokinase0.0637Diabetes5.12×10−25
8Platelets0.0498Anaemia6.21×10−25
9Smoking0.0139Platelets2.89×10−12
10Sex0.00410Serum sodium9.21×10−10
11Diabetes0.00211Ejection fraction7.22×10−09
   12Age5.34×10−05
  1. Results of the univariate application of the Pearson correlation coefficient between each feature and the target feature death event, absolute value (left), and the univariate application of the Shapiro–Wilk test on each feature (right)