Question

I am making a data mining application that can classify patients to their correct diagnosis, based on their symptoms. I was wondering if anyone knew what is the most suitable dataset I need and from where I can get it from. Is it a better approach to specialise in a certain area (ie Heart disease, diabetes etc) or should I classify patients on the whole, based on all types off illnesses? Ideally I want the dataset to be based on patients visits to their doctor. Therefore it should contain symptons and the diagnosis. However if I can get a dataset that contains other attributes such as age, gender etc then that would be even better. If anyone can help me and point me towards a suitable dataset then I would appreciate it.

Answer 1

There are some example datasets out there.

For example in Weka there is a dataset with diabetes data. These are the variables:

%    1. Number of times pregnant
%    2. Plasma glucose concentration a 2 hours in an oral glucose tolerance test
%    3. Diastolic blood pressure (mm Hg)
%    4. Triceps skin fold thickness (mm)
%    5. 2-Hour serum insulin (mu U/ml)
%    6. Body mass index (weight in kg/(height in m)^2)
%    7. Diabetes pedigree function
%    8. Age (years)
%    9. Class variable (0 or 1)

Also have a look at this page for a list of some more publicly available resources (some of which may have medical background):

• http://www.cs.waikato.ac.nz/ml/weka/datasets.html