Discrimination of indoor versus outdoor environmental state with machine learning algorithms in myopia observational studies

Table 1 Common classification type deep learning algorithms to determine positional accuracy

Machine learning algorithms	Accuracy % (N)
Gaussian process	78.4% (17,949/22,886)
Ensemble	79.7% (18,242/22,886)
Neural network	80.2% (18,361/22,886)
Discriminant analysis	83.8% (19,183/22,886)
Naive Bayes	87.4% (20,006/22,886)
Random forest	90.9% (20,805/22,886)
SVM	97.1% (22,229/22,886)
Total	85.4% (136,775/160,202)

All of the pairwise comparisons between these seven methods show significantly different (p < 0.001), except that between accuracy of neural network algorithm and average accuracy of these algorithms (p = 0.165)

ISSN: 1479-5876