Lilfeel commited on
Commit
94ce691
·
1 Parent(s): 83e9b93

Create winequality.names

Browse files
Files changed (1) hide show
  1. winequality.names +72 -0
winequality.names ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Citation Request:
2
+ This dataset is public available for research. The details are described in [Cortez et al., 2009].
3
+ Please include this citation if you plan to use this database:
4
+
5
+ P. Cortez, A. Cerdeira, F. Almeida, T. Matos and J. Reis.
6
+ Modeling wine preferences by data mining from physicochemical properties.
7
+ In Decision Support Systems, Elsevier, 47(4):547-553. ISSN: 0167-9236.
8
+
9
+ Available at: [@Elsevier] http://dx.doi.org/10.1016/j.dss.2009.05.016
10
+ [Pre-press (pdf)] http://www3.dsi.uminho.pt/pcortez/winequality09.pdf
11
+ [bib] http://www3.dsi.uminho.pt/pcortez/dss09.bib
12
+
13
+ 1. Title: Wine Quality
14
+
15
+ 2. Sources
16
+ Created by: Paulo Cortez (Univ. Minho), Antonio Cerdeira, Fernando Almeida, Telmo Matos and Jose Reis (CVRVV) @ 2009
17
+
18
+ 3. Past Usage:
19
+
20
+ P. Cortez, A. Cerdeira, F. Almeida, T. Matos and J. Reis.
21
+ Modeling wine preferences by data mining from physicochemical properties.
22
+ In Decision Support Systems, Elsevier, 47(4):547-553. ISSN: 0167-9236.
23
+
24
+ In the above reference, two datasets were created, using red and white wine samples.
25
+ The inputs include objective tests (e.g. PH values) and the output is based on sensory data
26
+ (median of at least 3 evaluations made by wine experts). Each expert graded the wine quality
27
+ between 0 (very bad) and 10 (very excellent). Several data mining methods were applied to model
28
+ these datasets under a regression approach. The support vector machine model achieved the
29
+ best results. Several metrics were computed: MAD, confusion matrix for a fixed error tolerance (T),
30
+ etc. Also, we plot the relative importances of the input variables (as measured by a sensitivity
31
+ analysis procedure).
32
+
33
+ 4. Relevant Information:
34
+
35
+ The two datasets are related to red and white variants of the Portuguese "Vinho Verde" wine.
36
+ For more details, consult: http://www.vinhoverde.pt/en/ or the reference [Cortez et al., 2009].
37
+ Due to privacy and logistic issues, only physicochemical (inputs) and sensory (the output) variables
38
+ are available (e.g. there is no data about grape types, wine brand, wine selling price, etc.).
39
+
40
+ These datasets can be viewed as classification or regression tasks.
41
+ The classes are ordered and not balanced (e.g. there are munch more normal wines than
42
+ excellent or poor ones). Outlier detection algorithms could be used to detect the few excellent
43
+ or poor wines. Also, we are not sure if all input variables are relevant. So
44
+ it could be interesting to test feature selection methods.
45
+
46
+ 5. Number of Instances: red wine - 1599; white wine - 4898.
47
+
48
+ 6. Number of Attributes: 11 + output attribute
49
+
50
+ Note: several of the attributes may be correlated, thus it makes sense to apply some sort of
51
+ feature selection.
52
+
53
+ 7. Attribute information:
54
+
55
+ For more information, read [Cortez et al., 2009].
56
+
57
+ Input variables (based on physicochemical tests):
58
+ 1 - fixed acidity
59
+ 2 - volatile acidity
60
+ 3 - citric acid
61
+ 4 - residual sugar
62
+ 5 - chlorides
63
+ 6 - free sulfur dioxide
64
+ 7 - total sulfur dioxide
65
+ 8 - density
66
+ 9 - pH
67
+ 10 - sulphates
68
+ 11 - alcohol
69
+ Output variable (based on sensory data):
70
+ 12 - quality (score between 0 and 10)
71
+
72
+ 8. Missing Attribute Values: None