1
$\begingroup$

I am working on the Webspam UK 2006 dataset.I have a dataset with around 98 different features. I am planning to apply Decision Tree C4.5 classifier on it. I have done feature selection using random forest and I am able to get the MeanDecreaseGini for each feature. Now my question is out of these 98 features how do I decide the total number of features to drop?

> importance(model_rf)
       MeanDecreaseGini
hostid        16.470372
HST_1          8.243852
HST_2          7.269966
HST_3          6.867545
HST_4         12.726805
HST_5         11.496738
HST_6         16.547676
HST_7         11.990072
HST_8         12.133017
HST_9         13.215381
HST_10        15.987938
HST_11         7.452778
HST_12         6.426637
HST_13         6.989566
HST_14         7.523594
HST_15         9.180248
HST_16        17.578608
HST_17        13.813119
HST_18        12.469869
HST_19         8.205692
HST_20         9.928898
HST_21         8.204766
HST_22        10.453883
HST_23         7.250751
HST_24         7.768237
HMG_25         6.888539
HMG_26         5.202057
HMG_27         7.272760
HMG_28         9.496095
HMG_29        22.814004
HMG_30         8.385849
HMG_31         8.962740
HMG_32         8.788728
HMG_33         8.500586
HMG_34         9.605438
HMG_35         6.916501
HMG_36         5.499630
HMG_37         7.707432
HMG_38         7.170538
HMG_39         7.519015
HMG_40        20.864126
HMG_41        13.891407
HMG_42        17.097263
HMG_43         5.907863
HMG_44         8.961656
HMG_45         6.505953
HMG_46         8.198271
HMG_47         6.683629
HMG_48         7.521221
AVG_49         6.978512
AVG_50        10.625453
AVG_51        11.787007
AVG_52        10.501896
AVG_53         8.998260
AVG_54         7.169787
AVG_55        19.950379
AVG_56        10.941224
AVG_57        11.911999
AVG_58        11.145681
AVG_59        23.529763
AVG_60         8.726543
AVG_61        11.881046
AVG_62         6.889028
AVG_63        12.834768
AVG_64         6.642953
AVG_65        16.349392
AVG_66         6.885495
AVG_67        15.492990
AVG_68         7.982041
AVG_69         7.274046
AVG_70         6.883014
AVG_71         6.684649
AVG_72         7.105827
STD_73         6.440758
STD_74         6.952049
STD_75         7.192637
STD_76         8.917629
STD_77        23.578747
STD_78         9.402747
STD_79        39.137215
STD_80        11.908751
STD_81        34.713739
STD_82         9.006674
STD_83        42.645293
STD_84         9.555861
STD_85        13.003373
STD_86         6.609194
STD_87        13.641984
STD_88         6.627306
STD_89         8.565469
STD_90         6.497847
STD_91         8.985288
STD_92         6.702219
STD_93         8.187475
STD_94        10.954584
STD_95         8.111179
STD_96         8.080897
$\endgroup$

closed as too broad by Aditya, Stephen Rauch, Siong Thye Goh, Toros91, Benji Albert May 15 '18 at 1:40

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.