Experiments

Intro ] [ discretizedtest1 ] [ discretizedtest2 ] [ normalized-discretizedtest1 ] [ normalized-discretizedtest2 ] [ normalizedtest1 ] [ normalizedtest2 ]

Up: Report ]

Naive Bayes Classifier

Class >50K: Prior probability = 0.24

age:  Normal Distribution. Mean = 0.3759 StandardDev = 0.1422 WeightSum = 7841 Precision = 0.013888888888888888
workclass:  Discrete Estimator. Counts =  4964 725 623 372 618 354 1 1  (Total = 7658)
fnlwgt:  Normal Distribution. Mean = 0.1193 StandardDev = 0.0696 WeightSum = 7841 Precision = 4.57854493841857E-5
education:  Discrete Estimator. Counts =  2222 1388 61 1676 424 266 362 28 41 34 960 7 63 307 17 1  (Total = 7857)
education-num:  Normal Distribution. Mean = 0.7074 StandardDev = 0.159 WeightSum = 7841 Precision = 0.06666666666666667
marital-status:  Discrete Estimator. Counts =  6693 464 492 67 86 35 11  (Total = 7848)
occupation:  Discrete Estimator. Counts =  284 930 138 984 1969 1860 87 251 508 116 321 2 212 2  (Total = 7664)
relationship:  Discrete Estimator. Counts =  746 68 5919 857 38 219  (Total = 7847)
race:  Discrete Estimator. Counts =  7118 277 37 26 388  (Total = 7846)
sex:  Discrete Estimator. Counts =  1180 6663  (Total = 7843)
capital-gain:  Normal Distribution. Mean = 0.0403 StandardDev = 0.1458 WeightSum = 7841 Precision = 0.00847457627118644
capital-loss:  Normal Distribution. Mean = 0.0447 StandardDev = 0.1364 WeightSum = 7841 Precision = 0.01098901098901099
hours-per-week:  Normal Distribution. Mean = 0.453 StandardDev = 0.1118 WeightSum = 7841 Precision = 0.010752688172043012
native-country:  Discrete Estimator. Counts =  7172 8 31 13 40 45 1 41 25 9 17 21 26 19 2 62 26 13 11 6 34 5 6 13 3 3 5 21 5 3 4 4 3 4 4 7 10 3 3 7 1  (Total = 7736)


Class <=50K: Prior probability = 0.76

age:  Normal Distribution. Mean = 0.2728 StandardDev = 0.1912 WeightSum = 24720 Precision = 0.013888888888888888
workclass:  Discrete Estimator. Counts =  17734 1818 495 590 1477 946 15 8  (Total = 23083)
fnlwgt:  Normal Distribution. Mean = 0.1209 StandardDev = 0.0723 WeightSum = 24720 Precision = 4.57854493841857E-5
education:  Discrete Estimator. Counts =  3135 5905 1116 8827 154 803 1022 488 607 401 765 163 872 108 318 52  (Total = 24736)
education-num:  Normal Distribution. Mean = 0.573 StandardDev = 0.1624 WeightSum = 24720 Precision = 0.06666666666666667
marital-status:  Discrete Estimator. Counts =  8285 3981 10193 960 909 385 14  (Total = 24727)
occupation:  Discrete Estimator. Counts =  646 3171 3159 2668 2099 2282 1285 1753 3264 880 1278 149 439 9  (Total = 23082)
relationship:  Discrete Estimator. Counts =  824 5002 7276 7450 945 3229  (Total = 24726)
race:  Discrete Estimator. Counts =  20700 764 276 247 2738  (Total = 24725)
sex:  Discrete Estimator. Counts =  9593 15129  (Total = 24722)
capital-gain:  Normal Distribution. Mean = 0.0015 StandardDev = 0.0097 WeightSum = 24720 Precision = 0.00847457627118644
capital-loss:  Normal Distribution. Mean = 0.0122 StandardDev = 0.0712 WeightSum = 24720 Precision = 0.01098901098901099
hours-per-week:  Normal Distribution. Mean = 0.3859 StandardDev = 0.1252 WeightSum = 24720 Precision = 0.010752688172043012
native-country:  Discrete Estimator. Counts =  22000 13 61 103 83 94 15 61 39 22 65 56 71 26 13 138 49 49 72 63 611 34 20 18 69 17 25 32 41 58 11 62 33 10 16 11 98 18 30 15 2  (Total = 24324)



=== Error on training data ===

Correctly Classified Instances       27152               83.3881 %
Incorrectly Classified Instances      5409               16.6119 %
Mean absolute error                      0.1742
Root mean squared error                  0.3729
Relative absolute error                 47.6412 %
Root relative squared error             87.2244 %
Total Number of Instances            32561     


=== Confusion Matrix ===

     a     b   <-- classified as
  4034  3807 |     a = >50K
  1602 23118 |     b = <=50K


=== Error on test data ===

Correctly Classified Instances       13509               82.974  %
Incorrectly Classified Instances      2772               17.026  %
Mean absolute error                      0.1763
Root mean squared error                  0.3758
Relative absolute error                 48.5315 %
Root relative squared error             88.4647 %
Total Number of Instances            16281     


=== Confusion Matrix ===

     a     b   <-- classified as
  1940  1906 |     a = >50K
   866 11569 |     b = <=50K

 

by: Keith A. Pray
Last Modified: July 4, 2004 8:59 AM
© 2004 - 1975 Keith A. Pray.
All rights reserved.