1. Title of Database: 
   Thyroid gland data. ('normal', hypo and hyper functioning)

2. Sources:
   (a) Original owner
   	Danny Coomans, Dept. of Maths. and Stats., James Cook University,
   	Townsville 4811, Australia.  email: madhc@manta.jcu.edu.au
   (b) Donor of database
	Stefan Aeberhard, Dept. of Comp. Science, James Cook University,
        Townsville 4811, Australia.  email: stefan@coral.cs.jcu.edu.au
   (c) Date received 
	July, 1992
 
3. Past Usage:
   - (a) Coomans, D., Broeckaert, M. Jonckheer M. and Massart D.L.,
       "Comparison of Multivariate Discriminant Techniques for 
	Clinical Data - Application to the Thyroid Functional State",
	Meth. Inform. Med. 22 (1983) pp. 93-101.

     (b) The data was used for comparing 16 different discriminant
	techniques, each trying to predict the state of the 
	thyroid gland.

   - (a) Coomans, D. and I. Broeckaert, "Potential Pattern Recognition
	in Cemical and Medical Decision Making", Research Studies Press,
	Letchworth, England.

     (b) Uses the data to compare different different kernel density 
	methods. Some achieve 100% correct classification.

4. Relevant Information Paragraph:

	Five lab. tests are used to try to predict whether a patient's
	thyroid to the class euthyroidism, hypothyroidism or
	hyperthyroidism. The diagnosis (the class label) was based
	on a complete medical record, including anamnesis, scan etc.
   
5. Number of Instances

	215

6. Number of Attributes 

	5

7. For Each Attribute: (please give both acronym and full name if both exist)

	Attr. no : 

	1:	Class attribute (1 = normal, 2 = hyper, 3 = hypo)
	2:	T3-resin uptake test. (A percentage)
	3:	Total Serum thyroxin as measured by the isotopic
		displacement method. 
	4: 	Total serum triiodothyronine as measured by radioimmuno
		assay.
	5: 	basal thyroid-stimulating hormone (TSH) as measured by 
		radioimmuno assay.
	6: 	Maximal absolute difference of TSH value after injection of 
		200 micro grams of thyrotropin-releasing hormone as compared
		to the basal value.

	All attributes are continuous.

	(Note, there is a slight possibility of having the 
	 attribute numbers mixed up, see [2a] if it matters.)


8. Missing Attribute Values: how many per each attribute?

	None 

9. Class Distribution: number of instances per class

	Class 1: (normal)	150
	Class 2: (hyper)	35
	Class 3: (hypo)		30