Table 10 Performance metrics of the six ML models trained on the hybrid synthetic dataset with selected features.

From: Machine learning-based models for screening of anemia and leukemia using features of complete blood count reports

Model

Accuracy

Classes

Precision

Recall

Specificity

Miss-rate

DT

0.947 ± 0.008

Normal

0.913 ± 0.021

0.967 ± 0.018

0.971 ± 0.008

0.033 ± 0.018

Anemia

0.957 ± 0.025

0.959 ± 0.022

0.987 ± 0.008

0.041 ± 0.022

Leukemia

0.945 ± 0.014

0.899 ± 0.040

0.984 ± 0.005

0.101 ± 0.040

Combination

0.971 ± 0.011

0.957 ± 0.016

0.987 ± 0.005

0.043 ± 0.016

RF

0.980 ± 0.007

Normal

0.975 ± 0.015

0.978 ± 0.014

0.992 ± 0.005

0.022 ± 0.014

Anemia

0.990 ± 0.008

0.986 ± 0.010

0.997 ± 0.002

0.014 ± 0.010

Leukemia

0.964 ± 0.019

0.968 ± 0.014

0.989 ± 0.006

0.032 ± 0.014

Combination

0.988 ± 0.006

0.986 ± 0.004

0.995 ± 0.002

0.014 ± 0.004

GBM

0.971 ± 0.008

Normal

0.938 ± 0.020

0.984 ± 0.007

0.979 ± 0.007

0.016 ± 0.007

Anemia

0.975 ± 0.012

0.988 ± 0.007

0.993 ± 0.004

0.012 ± 0.007

Leukemia

0.972 ± 0.010

0.937 ± 0.023

0.992 ± 0.003

0.063 ± 0.023

Combination

0.994 ± 0.003

0.973 ± 0.008

0.997 ± 0.001

0.027 ± 0.008

SVM

0.879 ± 0.021

Normal

0.830 ± 0.026

0.891 ± 0.025

0.942 ± 0.010

0.109 ± 0.025

Anemia

0.808 ± 0.047

0.983 ± 0.013

0.931 ± 0.020

0.017 ± 0.013

Leukemia

0.904 ± 0.031

0.746 ± 0.036

0.976 ± 0.008

0.254 ± 0.036

Combination

0.983 ± 0.009

0.893 ± 0.036

0.993 ± 0.004

0.107 ± 0.036

LR

0.816 ± 0.024

Normal

0.759 ± 0.043

0.827 ± 0.031

0.915 ± 0.022

0.173 ± 0.031

Anemia

0.727 ± 0.040

0.840 ± 0.037

0.907 ± 0.016

0.160 ± 0.037

Leukemia

0.836 ± 0.044

0.686 ± 0.012

0.959 ± 0.013

0.314 ± 0.012

Combination

0.942 ± 0.015

0.889 ± 0.045

0.976 ± 0.006

0.111 ± 0.045

MLP

0.879 ± 0.018

Normal

0.854 ± 0.051

0.844 ± 0.086

0.952 ± 0.023

0.156 ± 0.086

Anemia

0.801 ± 0.069

0.897 ± 0.069

0.931 ± 0.035

0.103 ± 0.069

Leukemia

0.898 ± 0.030

0.816 ± 0.050

0.972 ± 0.010

0.184 ± 0.050

Combination

0.970 ± 0.010

0.941 ± 0.021

0.987 ± 0.004

0.059 ± 0.021

Back to article page