I'm developing a data mining application for healthcare sector. I need a data set that would give information on age, blood pressure, specific gravity(of blood and plasma), sugar, amount of red blood cells, sodium level and some attributes like that sort of healthy people.
7
-
2Have you gone through the UCI Machine Learning Repository? (link). It seems that to get data on healthy people, you could use the control group from most any study.– philshemOct 2, 2016 at 9:10
-
Thank you for your thoughts. But yet can we assure that the control groups are from totally healthy people? For example in the control group of diabetes study,though they don't have diabetes can't there be people having kidney problems? I just think that way. Please correct me if I'm wrong.– JanodyaOct 2, 2016 at 15:26
-
specific gravity is a strange term in relation to humans/human health. What do you mean? (Please edit).– user4293Oct 4, 2016 at 13:23
-
@JanDoggen I have done the editing. Please take a look.– JanodyaOct 4, 2016 at 18:05
-
Please anybody having even a clue about a data set that would fit in, please help.– JanodyaOct 6, 2016 at 17:03
|
Show 2 more comments
1 Answer
I think this data set won't exist, because, while "unhealthy" can be often objectively defined, "healthy" is subjective. For example, a data set consisting of diagnostic data from a "healthy" person may in fact be measured on an not-yet-diagnosed "unhealthy" person, or a person that lists themselves as "healthy" but in fact has relatively high blood pressure for their age, ethnicity, demographic, etc. (With enough data, these points should show up as outliers.)
So, like I posted in my comment, I'd suggest finding an open data set of "unhealthy" people and then using the control groups, which are selected because they are not exhibiting symptoms. This is called "treatment and control groups."
Another thing: Similar to this question -
finding medical data for non-sick people isn't as easy as it sounds.
