3 ms·
Interesting. My PhD research is on precisely this topic-- nutrition diaries + machine learning for identifying the effect of foods on people with chronic illnes
by tansey 13y ago
Interesting. My PhD research is on precisely this topic-- nutrition diaries + machine learning for identifying the effect of foods on people with chronic illnesses, such as Crohn's or MS.
I think this site is a good idea. It'd be even better if you could add in amounts and somehow standardize the entries. Even better still if you could break foods down by ingredients.
Also, if you get a few users and want to have someone analyze your data in super cool Bayesian ways, email me!
- zen_boy 13y agoMy girlfriend has Ankylosing spondylitis. The exact cause of the continued inflammation is unknown. It has been hypothesised that the cause is a gut bacteria feeding off of gluten/starch/sugar/etc. We are super excited about trying out an elimination diet with Soylent, but knowing the real possibility of placebo and other factors, such as stress, we are trying to come up with a systematic way to measure the effect of Soylent diet on the wellbeing. Do you have any advice on how to go about conducting such a (food) experiment, what and how to track things and anything else to consider? Thanks!
- tansey 13y agoUnfortunately, it's pretty tough to do it as just one person. All of my models assume you have a population of patients that are tracking their diets. With one person, you probably are best off just tracking your foods, medications, and how you feel (numerically, say [0,10] or something). Then after a while you can do a regression based on how much of a given item you think you have in your system.