Synthetic Smart Home Dataset

This dataset is derived from the van Kasteren dataset (van Kasteren, T., Noulas, A., Englebienne, G., Kröse, B.: Accurate activity recognition in a home setting. In: Proceedings of the 10th International Conference on Ubiquitous Computing (UbiComp08), 2008) using a Markov model. The dataset has been validated with a conformance test, yielding positive results.

The dataset comprises three columns:

1. Day: Represents the day number, with each row corresponding to a specific day.
2. Activities: Represents the sequence of activities performed on a given day. The activities are listed in the order they occurred. Each activity is denoted by a number corresponding to a specific action:
   • 1: 'leave house'
   • 4: 'use toilet'
   • 5: 'take shower'
   • 10: 'go to bed'
   • 13: 'prepare breakfast'
   • 15: 'prepare dinner'
   • 17: 'get drink'
3. Durations: Represents the duration (in minutes) for each corresponding activity in the "Activities" column.

For example, on Day 1, the sequence of activities is recorded as [4, 4, 10, 4, 1, 4, 4, 10, 4], with corresponding durations [1, 1, 23.8, 1, 202.03, 1.81, 1, 25.86, 1.29]. This indicates that on Day 1, the individual started by spending 1 minute using the toilet, followed by another 1 minute in the toilet, then 23.8 minutes in bed, and so on.