dl_regression_single: Deep Learning Regression with Known Network Structure

In tianwei-zhang/easyAI: Democratize AI

Description

Deep Learning Regression with Known Network Structure

Usage

dl_regression_single(x, y, complexity, dropout, lr, validation_split, num_epoch,
  num_patience)

Arguments

x	training feature matrix
y	target matrix
complexity	a vector indicating numbers of hidden units in each layer, e.g. c(3,6,7) means 3 layers with 3, 6, 7 units in each layer
dropout	a vector indicating the dropout rate in each layer, e.g c(0.1,0.2,0.3)
lr	learning rate for the optimizer
validation_split	percent of data used for validation
num_epoch	number of epoches to go through during training
num_patience	number of patience in early stopping criteria

Value

returns a list object with three values: model: keras model contructed. A keras_model object loss: a vector containing loss value in each epoch accuracy: a vector containing accuracy value in each epoch

tianwei-zhang/easyAI documentation built on May 14, 2019, 12:48 p.m.