Table 3.
Non-zero coefficients for the one SE nested cross-validated lasso model
Variable | Coefficient |
unknown_1 | 2.48 |
public_1 | 1.62 |
private_1 | 1.43 |
perpetrator_syria | 1.38 |
perpetrator_russia_syria | 0.88 |
perpetrator_us_coalition | 0.24 |
(Intercept) | 0.11 |
Larger coefficients indicate that a variable stayed in the model for a longer time before being zeroed out by the penalisation and are thus more closely related to the outcome that healthcare infrastructure was attacked.