Extending numerical features
Numerical features can undergo various methods to create extended features from them. Previously, we saw how we can transform continuous numerical data into ordinal data. Now, we will dive into extending our numerical features further.
Before we go any deeper into these methods, we will introduce a new dataset to work with.
Activity recognition from the Single Chest-Mounted Accelerometer dataset
This dataset collects data from a wearable accelerometer, mounted on the chest, collected from fifteen participants performing seven activities. The sampling frequency of the accelerometer is 52 Hz and the accelerometer data is uncalibrated.
The dataset is separated by participant and contains the following:
- Sequential number
- x acceleration
- y acceleration
- z acceleration
- Label
Labels are codified by numbers and represent an activity, as follows:
- Working at a computer
- Standing up, walking, and going up/down stairs
- Standing
- Walking
- Going up/down stairs
- Walking and talking with someone...