predict_TSLA: Prediction from TSLA with new data

In TSLA: Tree-Guided Rare Feature Selection and Logic Aggregation

Prediction from TSLA with new data

Description

Generate prediction for the response.

Usage

predict_TSLA(
  object,
  X_1_new = NULL,
  X_2_new,
  type = c("response", "link"),
  ...
)

Arguments

object	A fit output from TSLA.fit().
X_1_new	New unpenalized features in matrix form.
X_2_new	New binary features in matrix form.
type	Two options: "response" or "link". The two options only differ for family="logit".
...	Other parameters.

Value

Predictions. The first dimension is indexed by the observation, the second dimension is indexed by \lambda, and the third dimension is indexed by \alpha.

TSLA documentation built on April 4, 2025, 2:14 a.m.