This is an extension of a previous analysis in which I visualized text sentiment associated with different pain treatments in a large online chronic pain forum. I wanted to use some machine learning to determine whether the commenters would get better over time based on their comment frequency on treatment types, times of their comments, etc. Below is a feature list of my data, with a quick description of each: chronic pain -- how often they commented in the 'chronic pain' forum lower back pain -- same for 'lower back pain' forum neck pain -- same for 'neck pain' forum upper back pain -- same for 'upper back pain' forum therapy -- how often they mentioned some sort of physical therapy, exercise, etc. in the first half of their comments over the counter -- comment frequency of counter drugs (acetaminophen, naproxen, ibuprofen, etc) opiod -- comment frequency of opiod drugs muscle relaxant -- comment frequency of muscle relaxant drugs be...
