Projects: Clinician In a Loop    Population Modeling    Indoor Air Quality

Clinician In a Loop (2017-2018)

Iterative Design of Visual Analytics for a Clinician-in-the-loop Smart Home: we design CIL, a clinician-in-the loop visual interface, that provides clinicians with patient behavior patterns to support remote health monitoring and assessment.
* funded by National Institues of Health (NIH)
* collaborated with clinicians, psychologist and nurses

My contribution was to calculate the below features to represent human behaviors based on smart home sensor data.
1. distribution of time spent in different area of the home.
2. daily sleep duration (daytime/nighttime/overall)
3. ovrall activity level in home
4. number of sleep interruptions
5. nighttime relative walking speed
6. baseline for comparisions between healthy situations and ailments showing up.
7. time-aligned location & activity distribution.

Visualization Demo


Population Modelling (2016-Present)
Our study identifies and models the patterns of human daily routines in 99 smart homes with diverse participants, and provides insights on behavior patterns and detection of deviations that indicate potential health problem. Not only can this lead to more effective medical interventions, but these findings may benefit other fields.
* funded by National Science Foundation (NSF)

Part 1: Data Collection and Processing (click a graph to zoom in)..
A list of 82 distributions that are used for the model fitting.
A complete list of home descriptive parameters.

Part 2: outlier detection and model fitting (click a graph to zoom in).
Part 2.1:
Modeling Fitting for the Personal Hygiene inter-event time based on all the smart home data.

Part 2.2:
outlier detection and model fitting for seven routine activities).
              ( seven activities of daily routine: Work, Sleep, Relax, Cook, Eat, Personal Hygiene, and Wash Dishes. )

Part 3: summarized results of modeling at population level and among subpopulations

Indoor Air Quality (2015-present)
Our research is about Integrated Measurements and Modeling Using US Smart Homes to Assess Climate Change and Human Behaviors Impact on Indoor Air Quality.
* granted by the Department of Energy and by the U.S. Environmental Protection Agency's Science to Achieve Results (STAR) program
* collaborated with people from civil and environmental engineering, atmospheric research lab, and design and construction

Part 1: data visualization and transformation: smarthome-based human behavior detection (click a graph to zoom in).
Part 2: finding the relationship between human behavior and indoor air quality based on three regression analyses and we report correlation coefficients that are moderate or large (r >= 0.3).
                                                                                             Summarized results
Part 3: feature extraction based on three learning algorithms.
                                                                             Results for aggregated dataset
                                                                             Results for IAQ1
                                                                             Results for IAQ2