Table 3.

Accuracies among text classifiers

	n=1	n=1–2
Multinomial naïve Bayes	Housing: 91.62% Tobacco: 70.87% Alcohol: 70.77%	Housing: 91.43% Tobacco: 77.18% Alcohol: 69.95%
Support vector machine	Housing: 92.18% Tobacco: 81.08% Alcohol: 76.50%	Housing: 91.99% Tobacco: 82.88% Alcohol: 81.97%
Logistic regression	Housing: 84.36% Tobacco: 75.38% Alcohol: 77.60%	Housing: 90.13% Tobacco: 84.68% Alcohol: 82.79%
Random forest	Housing: 90.50% Tobacco: 76.28% Alcohol: 71.31%	Housing: 91.25% Tobacco: 78.98% Alcohol: 75.68%

Bold values indicates highest performance for each SDBH.