152 cells were used for training, and 52 cells were used to test each algorithm (i) k-nearest neighbors, (ii) multinomial logistic regression, and (iii) XGBoost. The accuracy for the training set of 152 cells was 0.94, 0.98, and 1.00, respectively. The accuracy for the test set of 52 cells was 0.96, 0.96, and 0.92, respectively. Matthews Correlation Coefficients are 0.95, 0.95, and 0.91, respectively. 1=ionocytes, 2=tuft cells, 3=hillock cells, 4=ciliated cells, 5=secretory cells, 6=neuroendocrine cells, and 7=basal cells.