Table 1 Literature that uses AI and ML for smoking prediction.

²²

19,410 Instances

classification and regression trees

80% accuracy

-

²³

9 Instances

Neural Decision Forest (VARST) and a Variational Autoencoder (VAE)

96.29%.

F1-score

-

²⁴

55,693 Instances

Various supervised models.

84.73% accuracy

Implements LASSO for dimensionality reduction

²⁵

55,692 Instances

Various supervised models.

83.29%

accuracy

-

²⁶

991,346 Instances

Various supervised models.

79.65%

accuracy

Model transparency through XAI