Data
mfeat-karhunen

mfeat-karhunen

active ARFF Publicly available Visibility: public Uploaded 06-04-2014 by Jan van Rijn
0 likes downloaded by 0 people , 0 total downloads 0 issues 0 downvotes
  • study_14 study_1 study_170 study_661 study_249 study_697 study_509 study_407 study_781 study_117 study_122 study_213 study_118 study_251 study_391
Issue #Downvotes for this reason By


Loading wiki
Help us complete this description Edit
Author: Source: Unknown - Please cite: The multi-feature digit dataset ------------------------------- Oowned and donated by: ---------------------- Robert P.W. Duin Department of Applied Physics Delft University of Technology P.O. Box 5046, 2600 GA Delft The Netherlands email: duin@ph.tn.tudelft.nl http : //www.ph.tn.tudelft.nl/~duin tel +31 15 2786143 Usage ----- A slightly different version of the database is used in M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4, 1998, 381-386. M. van Breukelen and R.P.W. Duin, Neural Network Initialization by Combined Classifiers, in: A.K. Jain, S. Venkatesh, B.C. Lovell (eds.), ICPR'98, Proc. 14th Int. Conference on Pattern Recognition (Brisbane, Aug. 16-20), The database as it is is used in: A.K. Jain, R.P.W. Duin, J. Mao, Statisitcal Pattern Recognition: A Review, in preparation Description ----------- This dataset consists of features of handwritten numerals (`0'--`9') extracted from a collection of Dutch utility maps. 200 patterns per class (for a total of 2,000 patterns) have been digitized in binary images. These digits are represented in terms of the following six feature sets (files): 1. mfeat-fou: 76 Fourier coefficients of the character shapes; 2. mfeat-fac: 216 profile correlations; 3. mfeat-kar: 64 Karhunen-Love coefficients; 4. mfeat-pix: 240 pixel averages in 2 x 3 windows; 5. mfeat-zer: 47 Zernike moments; 6. mfeat-mor: 6 morphological features. In each file the 2000 patterns are stored in ASCI on 2000 lines. The first 200 patterns are of class `0', followed by sets of 200 patterns for each of the classes `1' - `9'. Corresponding patterns in different feature sets (files) correspond to the same original character. The source image dataset is lost. Using the pixel-dataset (mfeat-pix) sampled versions of the original images may be obtained (15 x 16 pixels). Total number of instances: -------------------------- 2000 (200 instances per class) Total number of attributes: --------------------------- 649 (distributed over 6 datasets,see above) no missing attributes Total number of classes: ------------------------ 10 Format: ------ 6 files, see above. Each file contains 2000 lines, one for each instance. Attributes are SPACE separated and can be loaded by Matlab as > load filename No missing attributes. Some are integer, others are real. Information about the dataset CLASSTYPE: nominal CLASSINDEX: last

65 features

class (target)nominal10 unique values
0 missing
att48numeric1994 unique values
0 missing
att34numeric1994 unique values
0 missing
att47numeric1994 unique values
0 missing
att46numeric1994 unique values
0 missing
att45numeric1994 unique values
0 missing
att44numeric1994 unique values
0 missing
att43numeric1994 unique values
0 missing
att42numeric1994 unique values
0 missing
att41numeric1993 unique values
0 missing
att40numeric1994 unique values
0 missing
att39numeric1994 unique values
0 missing
att38numeric1994 unique values
0 missing
att37numeric1994 unique values
0 missing
att36numeric1994 unique values
0 missing
att35numeric1994 unique values
0 missing
att33numeric1994 unique values
0 missing
att49numeric1994 unique values
0 missing
att50numeric1994 unique values
0 missing
att51numeric1994 unique values
0 missing
att52numeric1994 unique values
0 missing
att53numeric1994 unique values
0 missing
att54numeric1994 unique values
0 missing
att55numeric1994 unique values
0 missing
att56numeric1994 unique values
0 missing
att57numeric1994 unique values
0 missing
att58numeric1994 unique values
0 missing
att59numeric1994 unique values
0 missing
att60numeric1994 unique values
0 missing
att61numeric1994 unique values
0 missing
att62numeric1994 unique values
0 missing
att63numeric1993 unique values
0 missing
att64numeric1994 unique values
0 missing
att17numeric1994 unique values
0 missing
att2numeric1994 unique values
0 missing
att3numeric1994 unique values
0 missing
att4numeric1994 unique values
0 missing
att5numeric1994 unique values
0 missing
att6numeric1994 unique values
0 missing
att7numeric1994 unique values
0 missing
att8numeric1994 unique values
0 missing
att9numeric1994 unique values
0 missing
att10numeric1994 unique values
0 missing
att11numeric1994 unique values
0 missing
att12numeric1994 unique values
0 missing
att13numeric1994 unique values
0 missing
att14numeric1994 unique values
0 missing
att15numeric1994 unique values
0 missing
att16numeric1994 unique values
0 missing
att1numeric1994 unique values
0 missing
att18numeric1994 unique values
0 missing
att19numeric1994 unique values
0 missing
att20numeric1993 unique values
0 missing
att21numeric1994 unique values
0 missing
att22numeric1994 unique values
0 missing
att23numeric1994 unique values
0 missing
att24numeric1994 unique values
0 missing
att25numeric1994 unique values
0 missing
att26numeric1994 unique values
0 missing
att27numeric1994 unique values
0 missing
att28numeric1994 unique values
0 missing
att29numeric1994 unique values
0 missing
att30numeric1994 unique values
0 missing
att31numeric1994 unique values
0 missing
att32numeric1994 unique values
0 missing

107 properties

2000
Number of instances (rows) of the dataset.
65
Number of attributes (columns) of the dataset.
10
Number of distinct values of the target attribute (if it is nominal).
0
Number of missing values in the dataset.
0
Number of instances with at least one value missing.
64
Number of numeric attributes.
1
Number of nominal attributes.
1
Average class difference between consecutive instances.
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.21
Error rate achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.21
Error rate achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.21
Error rate achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
3.32
Entropy of the target attribute values.
0.69
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.DecisionStump
0.8
Error rate achieved by the landmarker weka.classifiers.trees.DecisionStump
0.11
Kappa coefficient achieved by the landmarker weka.classifiers.trees.DecisionStump
0.03
Number of attributes divided by the number of instances.
Number of attributes needed to optimally describe the class (under the assumption of independence among attributes). Equals ClassEntropy divided by MeanMutualInformation.
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .00001
0.21
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .00001
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .00001
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .0001
0.21
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .0001
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .0001
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .001
0.21
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .001
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .001
10
Percentage of instances belonging to the most frequent class.
200
Number of instances belonging to the most frequent class.
Maximum entropy among attributes.
0.58
Maximum kurtosis among attributes of the numeric type.
2.89
Maximum of means among attributes of the numeric type.
Maximum mutual information between the nominal attributes and the target attribute.
10
The maximum number of distinct values among attributes of the nominal type.
0.42
Maximum skewness among attributes of the numeric type.
7.69
Maximum standard deviation of attributes of the numeric type.
Average entropy of the attributes.
-0.05
Mean kurtosis among attributes of the numeric type.
0.05
Mean of means among attributes of the numeric type.
Average mutual information between the nominal attributes and the target attribute.
An estimate of the amount of irrelevant information in the attributes regarding the class. Equals (MeanAttributeEntropy - MeanMutualInformation) divided by MeanMutualInformation.
10
Average number of distinct values among the attributes of the nominal type.
-0
Mean skewness among attributes of the numeric type.
2.15
Mean standard deviation of attributes of the numeric type.
Minimal entropy among attributes.
-0.94
Minimum kurtosis among attributes of the numeric type.
-1.82
Minimum of means among attributes of the numeric type.
Minimal mutual information between the nominal attributes and the target attribute.
10
The minimal number of distinct values among attributes of the nominal type.
-0.34
Minimum skewness among attributes of the numeric type.
0.89
Minimum standard deviation of attributes of the numeric type.
10
Percentage of instances belonging to the least frequent class.
200
Number of instances belonging to the least frequent class.
1
Area Under the ROC Curve achieved by the landmarker weka.classifiers.bayes.NaiveBayes
0.07
Error rate achieved by the landmarker weka.classifiers.bayes.NaiveBayes
0.92
Kappa coefficient achieved by the landmarker weka.classifiers.bayes.NaiveBayes
0
Number of binary attributes.
0
Percentage of binary attributes.
0
Percentage of instances having missing values.
0
Percentage of missing values.
98.46
Percentage of numeric attributes.
1.54
Percentage of nominal attributes.
First quartile of entropy among attributes.
-0.15
First quartile of kurtosis among attributes of the numeric type.
-0.3
First quartile of means among attributes of the numeric type.
First quartile of mutual information between the nominal attributes and the target attribute.
-0.1
First quartile of skewness among attributes of the numeric type.
1.14
First quartile of standard deviation of attributes of the numeric type.
Second quartile (Median) of entropy among attributes.
-0.02
Second quartile (Median) of kurtosis among attributes of the numeric type.
0
Second quartile (Median) of means among attributes of the numeric type.
Second quartile (Median) of mutual information between the nominal attributes and the target attribute.
-0.02
Second quartile (Median) of skewness among attributes of the numeric type.
1.65
Second quartile (Median) of standard deviation of attributes of the numeric type.
Third quartile of entropy among attributes.
0.07
Third quartile of kurtosis among attributes of the numeric type.
0.33
Third quartile of means among attributes of the numeric type.
Third quartile of mutual information between the nominal attributes and the target attribute.
0.06
Third quartile of skewness among attributes of the numeric type.
2.66
Third quartile of standard deviation of attributes of the numeric type.
0.92
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.21
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.92
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 2
0.21
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 2
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 2
0.92
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.21
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.77
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.83
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
0.31
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
0.66
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
0.83
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.31
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.66
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.83
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0.31
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0.66
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0
Standard deviation of the number of distinct values among attributes of the nominal type.
0.97
Area Under the ROC Curve achieved by the landmarker weka.classifiers.lazy.IBk
0.05
Error rate achieved by the landmarker weka.classifiers.lazy.IBk
0.94
Kappa coefficient achieved by the landmarker weka.classifiers.lazy.IBk

11 tasks

0 runs - estimation_procedure: 10-fold Crossvalidation - target_feature: class
0 runs - estimation_procedure: 33% Holdout set - target_feature: class
0 runs - estimation_procedure: Test on Training Data - target_feature: class
0 runs - estimation_procedure: 20% Holdout (Ordered) - target_feature: class
0 runs - estimation_procedure: Leave one out - target_feature: class
0 runs - estimation_procedure: 5 times 2-fold Crossvalidation - target_feature: class
0 runs - estimation_procedure: 10 times 10-fold Crossvalidation - target_feature: class
0 runs - estimation_procedure: 10% Holdout set - target_feature: class
0 runs - estimation_procedure: 10 times 10-fold Learning Curve - target_feature: class
0 runs - estimation_procedure: 10-fold Learning Curve - target_feature: class
0 runs - estimation_procedure: Interleaved Test then Train - target_feature: class
Define a new task