fit: Search for the Best Model and Hyperparameters

In autokeras: R Interface to 'AutoKeras'

Description

It will search for the best model and hyperparameters based on the performances on validation data.

Usage

## S3 method for class 'AutokerasModel'
fit(
  object,
  x = NULL,
  y = NULL,
  epochs = 1000,
  callbacks = NULL,
  validation_split = 0.2,
  validation_data = NULL,
  ...
)

Arguments

object	: An AutokerasModel instance.
x	: Training data x. Check corresponding AutokerasModel help to note how it should be provided.
y	: Training data y. Check corresponding AutokerasModel help to note how it should be provided.
epochs	: numeric. The number of epochs to train each model during the search. If unspecified, by default we train for a maximum of '1000' epochs, but we stop training if the validation loss stops improving for 10 epochs (unless you specified an EarlyStopping callback as part of the 'callbacks' argument, in which case the EarlyStopping callback you specified will determine early stopping).
callbacks	: list of Keras callbacks to apply during training and validation.
validation_split	: numeric between 0 and 1. Defaults to '0.2'. Fraction of the training data to be used as validation data. The model will set apart this fraction of the training data, will not train on it, and will evaluate the loss and any model metrics on this data at the end of each epoch. The validation data is selected from the last samples in the 'x' and 'y' data provided, before shuffling. This argument is not supported when 'x' is a dataset. The best model found would be fit on the entire dataset including the validation data.
validation_data	: Data on which to evaluate the loss and any model metrics at the end of each epoch. The model will not be trained on this data. 'validation_data' will override 'validation_split'. The type of the validation data should be the same as the training data. The best model found would be fit on the training dataset without the validation data.
...	: Unused.

Value

A trained AutokerasModel.

Examples

## Not run: 
library("keras")

# use the MNIST dataset as an example
mnist <- dataset_mnist()
c(x_train, y_train) %<-% mnist$train
c(x_test, y_test) %<-% mnist$test

library("autokeras")

# Initialize the image classifier
clf <- model_image_classifier(max_trials = 10) %>% # It tries 10 different models
  fit(x_train, y_train) # Feed the image classifier with training data

# If you want to use own valitadion data do:
clf <- model_image_classifier(max_trials = 10) %>%
  fit(
    x_train,
    y_train,
    validation_data = list(x_test, y_test)
  )

# Predict with the best model
(predicted_y <- clf %>% predict(x_test))

# Evaluate the best model with testing data
clf %>% evaluate(x_test, y_test)

# Get the best trained Keras model, to work with the keras R library
export_model(clf)

## End(Not run)

autokeras documentation built on Jan. 17, 2021, 9:06 a.m.