gpb.grid.search.tune.parameters {gpboost}

Function for choosing tuning parameters

Description

Function that allows for choosing tuning parameters from a grid in a determinstic or random way using cross validation or validation data sets.

Usage

gpb.grid.search.tune.parameters(param_grid, data, params = list(),
  num_try_random = NULL, nrounds = 100L, gp_model = NULL,
  line_search_step_length = FALSE, use_gp_model_for_validation = TRUE,
  train_gp_model_cov_pars = TRUE, folds = NULL, nfold = 4L,
  label = NULL, weight = NULL, obj = NULL, eval = NULL,
  verbose_eval = 1L, stratified = TRUE, init_model = NULL,
  colnames = NULL, categorical_feature = NULL,
  early_stopping_rounds = NULL, callbacks = list(),
  return_all_combinations = FALSE, ...)

Arguments

param_grid

list with candidate parameters defining the grid over which a search is done

data

a gpb.Dataset object, used for training. Some functions, such as gpb.cv, may allow you to pass other types of data like matrix and then separately supply label as a keyword argument.

params

list with other parameters not included in param_grid

num_try_random

integer with number of random trial on parameter grid. If NULL, a deterministic search is done

nrounds

number of boosting iterations (= number of trees). This is the most important tuning parameter for boosting

gp_model

A GPModel object that contains the random effects (Gaussian process and / or grouped random effects) model

line_search_step_length

Boolean. If TRUE, a line search is done to find the optimal step length for every boosting update (see, e.g., Friedman 2001). This is then multiplied by the learning_rate. Applies only to the GPBoost algorithm

use_gp_model_for_validation

Boolean. If TRUE, the gp_model (Gaussian process and/or random effects) is also used (in addition to the tree model) for calculating predictions on the validation data. If FALSE, the gp_model (random effects part) is ignored for making predictions and only the tree ensemble is used for making predictions for calculating the validation / test error.

train_gp_model_cov_pars

Boolean. If TRUE, the covariance parameters of the gp_model (Gaussian process and/or random effects) are estimated in every boosting iterations, otherwise the gp_model parameters are not estimated. In the latter case, you need to either estimate them beforehand or provide the values via the init_cov_pars parameter when creating the gp_model

folds

list provides a possibility to use a list of pre-defined CV folds (each element must be a vector of test fold's indices). When folds are supplied, the nfold and stratified parameters are ignored.

nfold

the original dataset is randomly partitioned into nfold equal size subsamples.

label

Vector of labels, used if data is not an gpb.Dataset

weight

vector of response values. If not NULL, will set to dataset

obj

(character) The distribution of the response variable (=label) conditional on fixed and random effects. This only needs to be set when doing independent boosting without random effects / Gaussian processes.

eval

Evaluation metric to be monitored when doing CV and parameter tuning. This can be a string, function, or list with a mixture of strings and functions.

• a. character vector: Non-exhaustive list of supported metrics: "test_neg_log_likelihood", "mse", "rmse", "mae", "auc", "average_precision", "binary_logloss", "binary_error" See the "metric" section of the parameter documentation for a complete list of valid metrics.

• b. function: You can provide a custom evaluation function. This should accept the keyword arguments preds and dtrain and should return a named list with three elements:

  • name: A string with the name of the metric, used for printing and storing results.

  • value: A single number indicating the value of the metric for the given predictions and true values

  • higher_better: A boolean indicating whether higher values indicate a better fit. For example, this would be FALSE for metrics like MAE or RMSE.

• c. list: If a list is given, it should only contain character vectors and functions. These should follow the requirements from the descriptions above.

verbose_eval

integer. Whether to display information on the progress of tuning parameter choice. If None or 0, verbose is of. If = 1, summary progress information is displayed for every parameter combination. If >= 2, detailed progress is displayed at every boosting stage for every parameter combination.

stratified

a boolean indicating whether sampling of folds should be stratified by the values of outcome labels.

init_model

path of model file of gpb.Booster object, will continue training from this model

colnames

feature names, if not null, will use this to overwrite the names in dataset

categorical_feature

categorical features. This can either be a character vector of feature names or an integer vector with the indices of the features (e.g. c(1L, 10L) to say "the first and tenth columns").

early_stopping_rounds

int. Activates early stopping. Requires at least one validation data and one metric. When this parameter is non-null, training will stop if the evaluation of any metric on any validation set fails to improve for early_stopping_rounds consecutive boosting rounds. If training stops early, the returned model will have attribute best_iter set to the iteration number of the best iteration.

callbacks

List of callback functions that are applied at each iteration.

return_all_combinations

a boolean indicating whether all tried parameter combinations are returned

...

other parameters, see Parameters.rst for more information.

Value

A list with the best parameter combination and score The list has the following format: list("best_params" = best_params, "best_iter" = best_iter, "best_score" = best_score) If return_all_combinations is TRUE, then the list contains an additional entry 'all_combinations'

Early Stopping

"early stopping" refers to stopping the training process if the model's performance on a given validation set does not improve for several consecutive iterations.

If multiple arguments are given to eval, their order will be preserved. If you enable early stopping by setting early_stopping_rounds in params, by default all metrics will be considered for early stopping.

If you want to only consider the first metric for early stopping, pass first_metric_only = TRUE in params. Note that if you also specify metric in params, that metric will be considered the "first" one. If you omit metric, a default metric will be used based on your choice for the parameter obj (keyword argument) or objective (passed into params).

Author(s)

Fabio Sigrist

Examples

# See https://github.com/fabsig/GPBoost/tree/master/R-package for more examples

library(gpboost)
data(GPBoost_data, package = "gpboost")

# Create random effects model, dataset, and define parameter grid
gp_model <- GPModel(group_data = group_data[,1], likelihood="gaussian")
dataset <- gpb.Dataset(X, label = y)
param_grid = list("learning_rate" = c(1,0.1,0.01), 
                  "min_data_in_leaf" = c(10,100,1000),
                  "max_depth" = c(1,2,3,5,10),
                  "lambda_l2" = c(0,1,10))
other_params <- list(num_leaves = 2^10)
# Note: here we try different values for 'max_depth' and thus set 'num_leaves' to a large value.
#       An alternative strategy is to impose no limit on 'max_depth', 
#       and try different values for 'num_leaves' as follows:
# param_grid = list("learning_rate" = c(1,0.1,0.01), 
#                   "min_data_in_leaf" = c(10,100,1000),
#                   "num_leaves" = 2^(1:10),
#                   "lambda_l2" = c(0,1,10))
# other_params <- list(max_depth = -1)
set.seed(1)
opt_params <- gpb.grid.search.tune.parameters(param_grid = param_grid, params = other_params,
                                              num_try_random = NULL, nfold = 4,
                                              data = dataset, gp_model = gp_model,
                                              use_gp_model_for_validation=TRUE, verbose_eval = 1,
                                              nrounds = 1000, early_stopping_rounds = 10)
print(paste0("Best parameters: ",
             paste0(unlist(lapply(seq_along(opt_params$best_params), 
                                  function(y, n, i) { paste0(n[[i]],": ", y[[i]]) }, 
                                  y=opt_params$best_params, 
                                  n=names(opt_params$best_params))), collapse=", ")))
print(paste0("Best number of iterations: ", opt_params$best_iter))
print(paste0("Best score: ", round(opt_params$best_score, digits=3)))
# Note: other scoring / evaluation metrics can be chosen using the 
#       'metric' argument, e.g., metric = "l1"

# Using manually defined validation data instead of cross-validation
valid_tune_idx <- sample.int(length(y), as.integer(0.2*length(y)))
folds = list(valid_tune_idx)
opt_params <- gpb.grid.search.tune.parameters(param_grid = param_grid, params = other_params,
                                              num_try_random = NULL, folds = folds,
                                              data = dataset, gp_model = gp_model,
                                              use_gp_model_for_validation=TRUE, verbose_eval = 1,
                                              nrounds = 1000, early_stopping_rounds = 10)