Bayesian Regression Analysis for Dependent Data with an Elliptical Shape
Author Information
Author(s): Yu Yian, Tang Long, Ren Kang, Chen Zhonglue, Chen Shengdi, Shi Jianqing
Primary Institution: Southern University of Science and Technology
Hypothesis
Can a parametric hierarchical model effectively capture dependencies in functional data with an elliptical shape?
Conclusion
The proposed model successfully reconstructs curved trajectories and captures systematic errors in gait data.
Supporting Evidence
- The model effectively captures the dynamics of gait patterns in patients with Parkinson's disease.
- Elliptical features derived from the model can aid in detecting freezing of gait.
- The proposed method requires fewer sensors compared to traditional approaches.
Takeaway
This study created a new way to analyze walking patterns in people with Parkinson's disease by fitting curves to their movement data.
Methodology
The study used a Bayesian hierarchical model with a Gaussian process prior and MCMC methods for inference.
Potential Biases
Potential biases may arise from the assumptions of independence in the model.
Limitations
The model primarily focuses on two-dimensional data and may require adjustments for higher-dimensional applications.
Participant Demographics
Participants included individuals diagnosed with dementia or Parkinson's disease.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website