Now that we have covered the basics, let's look at building our first true deep learning model! We will use the UHI HAR dataset that we used in Chapter 2, Training a Prediction Model. The following code does some data preparation: it loads the data and selects only the columns that store mean values (those that have the word mean in the column name). The y variables are from 1 to 6; we will subtract one so that the range is 0 to 5. The code for this section is in Chapter4/uci_har.R. It requires the UHI HAR dataset to be in the data folder; download it from https://archive.ics.uci.edu/ml/datasets/human+activity+recognition+using+smartphones and unzip it into the data folder:
train.x <- read.table("../data/UCI ...
