Table 7 The setup of the baseline machine learning models
From: Toward multi-label sentiment analysis: a transfer learning based approach
Base classifier | Hyper-parameters |
|---|---|
SGD | Loss = ‘hinge’, penalty = ‘l2’, \(\alpha\) = \(\frac{1}{training\_size \times 5}\), maximal iteration = 20, tolerance = \(1 \times 10^{-3}\) |
LR | Solver = ‘lbfgs’ |
RF | Num of trees = 100, maximal depth = 3 |