Deep Learning Based on R Language - Simple Regression Case

This article mainly addresses: How to use neural networks to build models?, other tweets can be seen: Deep Learning Based on R Language - Configuration Environment; Deep Learning Based on R Language - Introduction and Information Sharing.

This case is under CPU.If your device has a GPU and want to use the GPU to train the model.You don't need to modify the following code, just install the GPU version of TensorFlow in advance. By default, the operation will use the GPU first.

Knowledge points include:

Data import and data processing.
Building Neural Networks.
Training Neural Networks.
Evaluate the accuracy of the model.
Save and restore created models.
Loading Package

library(keras)library(mlbench) #using internal datalibrary(dplyr)library(magrittr)

Loading Data
Use housing data for Boston's 506 census tracts in 1970 as an example.The dataset has a total of 14 columns and 506 rows.where the dependent variable is medv (median asking price for owner-occupied housing, in $1,000), and the independent variables are 13 other variables, including ࿱