Table 1 The mean and standard deviation balance accuracy score (with respect to 30 datasets) for the best performing image feature and classifiers (up to and including HQC and classifiers where the score was tied with HQC), with corresponding mean and standard deviation Jaccard index and Dice coefficient, for cell lines MDA-MD-231, U87-MG, MCF7 and U251.
From: A quantum-inspired classifier for clonogenic assay evaluations
Cell line | Best image feature | Best classifier | Balanced accuracy | Jaccard index | Dice coefficient |
|---|---|---|---|---|---|
MDA-MD-231 | Homogeneity | Helstrom Quantum Classifier | 0.959 ± 0.036 | 0.918 ± 0.080 | 0.955 ± 0.046 |
U87-MG | Homogeneity | Helstrom Quantum Classifier | 0.919 ± 0.050 | 0.790 ± 0.121 | 0.877 ± 0.078 |
MCF7 | L*u*v* | Gaussian Naïve Bayes | 0.969 ± 0.034 | 0.892 ± 0.088 | 0.941 ± 0.050 |
L*u*v* | Helstrom Quantum Classifier | 0.965 ± 0.033 | 0.882 ± 0.084 | 0.935 ± 0.048 | |
L*u*v* | Multi Layer Perceptron | 0.965 ± 0.042 | 0.898 ± 0.100 | 0.943 ± 0.059 | |
U251 | Homogeneity | SVM - RBF | 0.980 ± 0.033 | 0.948 ± 0.079 | 0.971 ± 0.047 |
Homogeneity | Helstrom Quantum Classifier | 0.979 ± 0.029 | 0.944 ± 0.078 | 0.970 ± 0.045 |
