我尝试使用h2o的交叉验证。指标,但是我发现标准差与我计算的结果不同
mean sd cv_1_valid cv_2_valid cv_3_valid cv_4_valid cv_5_valid model
auc 0.9512599 0.0058884337 0.96505374 0.9521365 0.9406109 0.9451134 0.95338506 drf
sd(c(0.96505374,0.9521365,0.9406109,0.9451134,0.95338506) )
[1] 0.009310416
mean(c(0.96505374,0.9521365,0.9406109,0.9451134,0.95338506) )
[1] 0.9512599
平均值相同,但sd不同,还有其他计算方法吗?