R/bayesian_optimisation.R
BayesianOptimization.RdBayesian optimization oracle.
BayesianOptimization(
<<<<<<< HEAD
objective,
max_trials,
=======
objective = NULL,
max_trials = 10,
>>>>>>> e5e5fc418dff9176842e75a405470546a3342c42
num_initial_points = NULL,
alpha = 1e-04,
beta = 2.6,
seed = NULL,
hyperparameters = NULL,
allow_new_entries = TRUE,
<<<<<<< HEAD
tune_new_entries = TRUE
)String or `kerastuner.Objective`. If a string, the direction of the optimization (min or max) will be inferred.
Int. Total number of trials (model configurations) to test at most. Note that the oracle may interrupt the search before `max_trial` models have been tested if the search space has been exhausted.
(Optional) Int. The number of randomly generated samples as initial training data for Bayesian optimization. If not specified, a value of 3 times the dimensionality of the hyperparameter space is used.
Float. Value added to the diagonal of the kernel matrix during fitting. It represents the expected amount of noise in the observed performances in Bayesian optimization.
Float. The balancing factor of exploration and exploitation. The larger it is, the more explorative it is.
Int. Random seed.
HyperParameters class instance. Can be used to override (or register in advance) hyperparamters in the search space.
Whether the hypermodel is allowed to request hyperparameter entries not listed in `hyperparameters`.
Whether hyperparameter entries that are requested by the hypermodel but that were not specified in `hyperparameters` should be added to the search space, or not. If not, then the default value for these parameters will be used.
A string, `keras_tuner.Objective` instance, or a list of `keras_tuner.Objective`s and strings. If a string, the direction of the optimization (min or max) will be inferred. If a list of `keras_tuner.Objective`, we will minimize the sum of all the objectives to minimize subtracting the sum of all the objectives to maximize. The `objective` argument is optional when `Tuner.run_trial()` or `HyperModel.fit()` returns a single float as the objective to minimize.
Integer, the total number of trials (model configurations) to test at most. Note that the oracle may interrupt the search before `max_trial` models have been tested if the search space has been exhausted. Defaults to 10.
Optional number of randomly generated samples as initial training data for Bayesian optimization. If left unspecified, a value of 3 times the dimensionality of the hyperparameter space is used.
Float, the value added to the diagonal of the kernel matrix during fitting. It represents the expected amount of noise in the observed performances in Bayesian optimization. Defaults to 1e-4.
Float, the balancing factor of exploration and exploitation. The larger it is, the more explorative it is. Defaults to 2.6.
Optional integer, the random seed.
Optional `HyperParameters` instance. Can be used to override (or register in advance) hyperparameters in the search space.
Boolean, whether the hypermodel is allowed to request hyperparameter entries not listed in `hyperparameters`. Defaults to TRUE.
Boolean, whether hyperparameter entries that are requested by the hypermodel but that were not specified in `hyperparameters` should be added to the search space, or not. If not, then the default value for these parameters will be used. Defaults to TRUE.
Integer. Defaults to 0. The maximum number of times to retry a `Trial` if the trial crashed or the results are invalid.
Integer. Defaults to 3. The maximum number of consecutive failed `Trial`s. When this number is reached, the search will be stopped. A `Trial` is marked as failed when none of the retries succeeded.
BayesianOptimization tuning with Gaussian process
It uses Bayesian optimization with a underlying Gaussian process model. The acquisition function used is upper confidence bound (UCB), which can be found in the following link: https://www.cse.wustl.edu/~garnett/cse515t/spring_2015/files/lecture_notes/12.pdf
https://www.cse.wustl.edu/~garnett/cse515t/spring_2015/files/lecture_notes/12.pdf
=======BayesianOptimization tuning with Gaussian process
It uses Bayesian optimization with a underlying Gaussian process model. The acquisition function used is upper confidence bound (UCB), which can be found [here]( https://www.cse.wustl.edu/~garnett/cse515t/spring_2015/files/lecture_notes/12.pdf).
>>>>>>> e5e5fc418dff9176842e75a405470546a3342c42
if (FALSE) {
# The usage of 'tf$keras'
library(tensorflow)
tf$keras$Input(shape=list(28L, 28L, 1L))
}