Mining Human
Activity Patterns from Smart Home Big Data for Health Care Applications
|
City: |
London,
England |
|
Organization: |
Department
of Software Engineering, Lakehead University; Department of Electrical
Engineering and Computer Science, University of Ottawa; and Department of
Software Engineering, King Saud University |
|
Project
Start Date: |
1
January 2012 |
|
Project
End Date: |
31 December
2015 |
|
Reference: |
Yassine, A., Singh, S., & Alamri, A. (2017). Mining human activity patterns from smart
home big data for health care applications. IEEE Access, 5,
13131-13141. |
|
Problem: |
The
population in urban areas is growing continuously. As a result, the demand
for healthcare resources is also increasing over time. The change in the demographics
of cities places an enormous burden on cities to provide health services to
residents. In order to allocate resources efficiently, it is important to
track people’s habits and identify everyday routines in order to recognize
anomalous activities that indicate changes in health conditions. This will
help cities allocate resources more efficiently, in real-time, towards
helping individuals who need medical attention, such as the elderly or
individuals who live alone. |
|
Technical
Solution: |
· Clustering Analysis was used to
discover appliance-to-time associations. In other words, the method was used
to determine which appliances were used at different hours of the day. The
method was also used to determine appliance-to-appliance associations. Essentially,
they created clusters to determine which appliances in the house were used at
the same time. This helped build profiles about each household’s routine. · Bayesian Network was used to
predict short-term and long-term activities based on individual and multiple
appliance usage |
|
Datasets
Used: |
K. Jack and K.William, ``The UK-DALE
dataset, domestic appliance-level electricity demand and whole-house demand
from five UK homes,'' Sci. Data, vol. 2, p. 150007, Sep. 2015. |
|
Outcome: |
Pre-Solution Performance:
unknown; Post Solution Performance: short term accuracy for houses 1, 2, 3,
4, and 5 is 92.31%, 100.00%, 66.67%, 100.00%, and 100.00% respectively. The
obtained long-term accuracy for houses 1, 2, 3, 4, and 5 is 90.91%, 90.00%,
70.00%, 70.00%, and 80.00% respectively |
|
Issues
that arose: |
Not
applicable |
|
Status: |
In
Development |
|
Entered
by: |
28
September 2019: Sana Maqbool, Sana.Maqbool@mail.utoronto.ca |
CEM1002,
Civil Engineering, University of Toronto
Contact: msf@eil.utoronto.ca