I investigated, using a convolutional neural network (CNN), the classification of simulation data from a human aging model to predict mortality. I worked on this project part-time with theoretical physicist Dr. Andrew Rutenberg. This repo was made public on July 29, 2019.
The execution of CNN simulations was made possible by Compute Canada's Graham and Cedar clusters.
Question: Can the CNN predict whether a model individual will reach mortality in 5 years at age 80? Mortality in 10 years? Question: Does the accurcy of the prediction improve given more history of individuals' health trajectory?
Question: How does accuracy change with different configurations?
