parsnip (version 0.1.1)

mlp: General Interface for Single Layer Neural Network

Description

mlp(), for multilayer perceptron, is a way to generate a specification of a model before fitting and allows the model to be created using different packages in R or via keras The main arguments for the model are:

  • hidden_units: The number of units in the hidden layer (default: 5).

  • penalty: The amount of L2 regularization (aka weight decay, default is zero).

  • dropout: The proportion of parameters randomly dropped out of the model (keras only, default is zero).

  • epochs: The number of training iterations (default: 20).

  • activation: The type of function that connects the hidden layer and the input variables (keras only, default is softmax).

If parameters need to be modified, this function can be used in lieu of recreating the object from scratch.

Usage

mlp(
  mode = "unknown",
  hidden_units = NULL,
  penalty = NULL,
  dropout = NULL,
  epochs = NULL,
  activation = NULL
)

# S3 method for mlp update( object, parameters = NULL, hidden_units = NULL, penalty = NULL, dropout = NULL, epochs = NULL, activation = NULL, fresh = FALSE, ... )

Arguments

mode

A single character string for the type of model. Possible values for this model are "unknown", "regression", or "classification".

hidden_units

An integer for the number of units in the hidden model.

penalty

A non-negative numeric value for the amount of weight decay.

dropout

A number between 0 (inclusive) and 1 denoting the proportion of model parameters randomly set to zero during model training.

epochs

An integer for the number of training iterations.

activation

A single character strong denoting the type of relationship between the original predictors and the hidden unit layer. The activation function between the hidden and output layers is automatically set to either "linear" or "softmax" depending on the type of outcome. Possible values are: "linear", "softmax", "relu", and "elu"

object

A random forest model specification.

parameters

A 1-row tibble or named list with main parameters to update. If the individual arguments are used, these will supersede the values in parameters. Also, using engine arguments in this object will result in an error.

fresh

A logical for whether the arguments should be modified in-place of or replaced wholesale.

...

Not used for update().

Engine Details

Engines may have pre-set default arguments when executing the model fit call. For this type of model, the template of the fit calls are below:

keras

mlp() %>% 
  set_engine("keras") %>% 
  set_mode("regression") %>% 
  translate()
## Single Layer Neural Network Specification (regression)
## 
## Computational engine: keras 
## 
## Model fit template:
## parsnip::keras_mlp(x = missing_arg(), y = missing_arg())
mlp() %>% 
  set_engine("keras") %>% 
  set_mode("classification") %>% 
  translate()
## Single Layer Neural Network Specification (classification)
## 
## Computational engine: keras 
## 
## Model fit template:
## parsnip::keras_mlp(x = missing_arg(), y = missing_arg())

An error is thrown if both penalty and dropout are specified for keras models.

nnet

mlp() %>% 
  set_engine("nnet") %>% 
  set_mode("regression") %>% 
  translate()
## Single Layer Neural Network Specification (regression)
## 
## Main Arguments:
##   hidden_units = 5
## 
## Computational engine: nnet 
## 
## Model fit template:
## nnet::nnet(formula = missing_arg(), data = missing_arg(), weights = missing_arg(), 
##     size = 5, trace = FALSE, linout = TRUE)
mlp() %>% 
  set_engine("nnet") %>% 
  set_mode("classification") %>% 
  translate()
## Single Layer Neural Network Specification (classification)
## 
## Main Arguments:
##   hidden_units = 5
## 
## Computational engine: nnet 
## 
## Model fit template:
## nnet::nnet(formula = missing_arg(), data = missing_arg(), weights = missing_arg(), 
##     size = 5, trace = FALSE, linout = FALSE)

Parameter translations

The standardized parameter names in parsnip can be mapped to their original names in each engine that has main parameters:

parsnip keras nnet
hidden_units hidden_units size
penalty penalty decay
dropout dropout NA
epochs epochs maxit
activation activation NA

Details

These arguments are converted to their specific names at the time that the model is fit. Other options and argument can be set using set_engine(). If left to their defaults here (see above), the values are taken from the underlying model functions. One exception is hidden_units when nnet::nnet is used; that function's size argument has no default so a value of 5 units will be used. Also, unless otherwise specified, the linout argument to nnet::nnet() will be set to TRUE when a regression model is created. If parameters need to be modified, update() can be used in lieu of recreating the object from scratch.

The model can be created using the fit() function using the following engines:

  • R: "nnet" (the default)

  • keras: "keras"

See Also

fit()

Examples

Run this code
# NOT RUN {
mlp(mode = "classification", penalty = 0.01)
# Parameters can be represented by a placeholder:
mlp(mode = "regression", hidden_units = varying())
model <- mlp(hidden_units = 10, dropout = 0.30)
model
update(model, hidden_units = 2)
update(model, hidden_units = 2, fresh = TRUE)
# }

Run the code above in your browser using DataCamp Workspace