Table 11. Machine learning algorithms with one hot encoding and PCA using Python.
Race Group | Sample Size | Classification Accuracy (%) | |||
---|---|---|---|---|---|
BoW-PCA-SVM | BoW-PCA-RF | BoW-PCA-LDA | BoW-PCA-QDA | ||
AFRICAN | 90 | 84.2 | 85 | 88.2 | 64 |
ASIAN | 90 | 93.3 | 100 | 88.2 | 93.3 |
CAUCASIAN | 90 | 100 | 100 | 100 | 100 |
Micro-accuracy | 95.17 | 97.89 | 94.80 | 88.44 | |
Macro-accuracy | 92.50 | 95.00 | 92.13 | 85.77 |