1  1. Title: Miba Friction Plate Testing Data


2  2. Sources:


3  (a) Miba frictec, Andreas Promberger, Peter Mitterbauer Str. 1,


4  A4661 Roitham, AUSTRIA, +43 76139020, Andreas.promberger@miba.com


5  (b) FH Upper Austria, Gabriel Kronberger, Softwarepark 11,


6  A4232 Hagenberg, AUSTRIA, +43 50804 22320,


7  gabriel.kronberger@heuristiclab.com


8  (c) April 2017


9 


10 


11  3. Past Usage:


12   G. Kronberger, M. Kommenda, E. Lughofer, S. SamingerPlatz,


13  A. Promberger, F. Nickel, S. Winkler, M. Affenzeller  Robust


14  Generalized Fuzzy Modeling and Enhanced Symbolic Regression for


15  Modeling Tribological Systems, submitted to Applied Soft


16  Computing, 2017


17 


18   E. Lughofer, G. Kronberger, M. Kommenda, S. SamingerPlatz,


19  A. Promberger, F. Nickel, S. M. Winkler, M. Affenzeller  Robust


20  Fuzzy Modeling and Symbolic Regression for Establishing Accurate


21  and Interpretable Prediction Models in Supervising Tribological


22  Systems  Proceedings of the 8th International Joint Conference on


23  Computational Intelligence, Porto, Portugal, 2016, pp. 5163


24 


25  4. Relevant information:


26  A set of datasets for regression modelling of


27  friction characteristics of friction plate systems. The data stem


28  from tests of friction plates with commercial test benches for wet


29  friction plate systems. Friction characteristics such as the


30  coefficient of friction, wear, and temperatures are measured at


31  different loads. The goal is to predict these values given load


32  parameters. Data have been kindly provided by Miba frictec company.


33 


34  5. Number of instances


35  A separate file is provided for each target variable. The values of


36  the target variable are given in the last column.


37 


38   Cf1: 815 instances


39   Cf2: 2921 instances


40   Cf3: 657 instances


41   Cf4: 649 instances


42   NvhRating: 3943 instances


43   Temp1: 656 instances


44   Temp2: 648 instances


45   Wear1: 904 instances


46   Wear2: 902 instances


47 


48  6. Number of Attributes: 28


49  (2 binary, 22 numeric & continuous, and four nominal) Depending on


50  the target variable (or file) some of the attributes might be


51  constant.


52 


53  7. Attribute Information


54  The first column 'Partition' contains the partition assignment that


55  should be used for validation of algorithms. This attribute should


56  not be used for modelling. Rows must not be shuffled as in standard


57  crossvalidation.


58 


59   Source1: indicates that the row stems from data source one [binary]


60   Source2: indicates that the row stems from data source two (


61  redundant given Source1) [binary]


62   x1: an attribute of the testing procedure [numeric, integer]


63   Material_Cat: represents the type of friction material [nominal]


64   x2, ... ,x16: material attributes [numeric, continuous]


65   Material: represents the friction material [nominal]


66   Grooving: represents the grooving (surface structure) on the


67  friction plate [nominal]


68   Oil: represents the oil type [nominal]


69   x17, ... ,x22 load attributes [numeric, continuous]


70 


71 


72  If you use these data files please use the following reference:


73  G. Kronberger, M. Kommenda, E. Lughofer, S. SamingerPlatz,


74  A. Promberger, F. Nickel, S. Winkler, M. Affenzeller  Robust


75  Generalized Fuzzy Modeling and Enhanced Symbolic Regression for


76  Modeling Tribological Systems, submitted to Applied Soft Computing,


77  2017


78 

