Fit a gradient boosting regression model with XGBoost

Fits a gradient boosted tree model using the XGBoost implementation.

fit_xgb(
  formula,
  data,
  params = list(max_depth = 3, eta = 0.01, objective = "reg:squarederror"),
  nrounds = 1000,
  nthread = -1,
  ...
)

Arguments

formula

An object of class formula with a symbolic description of the model to be fitted.

data

A data.frame containing the variables in the model.

params

A list of XGBoost parameters including:

• max_depth - Maximum tree depth

• eta - Learning rate

• objective - Loss function to optimize

nrounds

Maximum number of boosting iterations.

nthread

Number of parallel threads used (-1 for all cores).

...

Additional parameters passed to xgboost::xgboost.

Value

A list containing two data frames:

• metrics - Model performance metrics

• coefficients - Feature importance measures