Generalized Linear Models of home activity for automatic detection of mild cognitive impairment in older adults.

摘要: With a globally aging population, the burden of care cognitively impaired older adults is becoming increas- ingly concerning. Instances Alzheimer's disease and other forms dementia are ever more frequent. Earlier detection cognitive impairment offers significant benefits, but remains difficult to do in practice. In this paper, we develop statistical models behavior within their homes using sensor data order detect early onset decline. Specifically, use inhomogenous Poisson processes model presence subjects different rooms throughout day home unobtrusive sensing technologies. We compare distributions learned from intact information theoretic tools observe differences between two populations which believe can be used help