Abstract
Survival analysis can be used to estimate the lifespan of a particular population that is being studied. It can be used in a large number of fields that include, but are not limited to medicine, public health, biology, engineering, and marketing. Depending on if the individual has experienced the event of interest or not, the data can be censored or uncensored. It is important to be aware of this when conducting analysis. The Kaplan-Meier estimate incorporates this and shows the probability of surviving in a given length of time over many small time intervals. The data set used contains 301 individuals between the ages of sixty-five and ninety-six years old to predict the time until their first and second falls. Information such as walking speed, stride length, use of assistive devices, previous falls, and gate smoothness were used as explanatory variables. LASSO feature selection was used to reduce the dimensionality of the data set and reduce the amount of non-significant variables from the prediction. Once the variables were selected, a cox proportional hazard model was conducted.
How to Cite
Brooks, L. & Dziergas, D., (2020) “Predicting Falls in Older People with the LASSO Method”, Capstone, The UNC Asheville Journal of Undergraduate Scholarship 33(1).
3
Views
1
Downloads