{{category R}}
{{category R.package}}
!!!randomForest ランダムフォレスト
RandomForest
{{pre
library(randomForest)
randomForest 4.6-14
Type rfNews() to see new features/changes/bug fixes.
 
 
Score
<fctr>
Token
<int>
Type
<int>
NoS
<int>
TTR
<dbl>
GI
<dbl>
MATTR
<dbl>
AWL
<dbl>
ASL
<dbl>
1	4	319	135	30	0.4231975	7.558549	0.5921317	4.304075	10.63333
2	4	356	161	29	0.4522472	8.532983	0.6649157	4.233146	12.27586
3	3	201	121	13	0.6019900	8.534682	0.7170149	4.746269	15.46154
4	4	260	140	27	0.5384615	8.682431	0.6877692	4.761538	9.62963
5	4	420	175	25	0.4166667	8.539126	0.6341905	3.995238	16.80000
6	3	261	124	20	0.4750958	7.675407	0.6390038	4.072797	13.05000
6 rows
jpn.RFmodel

Call:
 randomForest(formula = Score ~ ., data = jpn.5c) 
               Type of random forest: classification
                     Number of trees: 500
No. of variables tried at each split: 2

        OOB estimate of  error rate: 17.19%
Confusion matrix:
  1 2   3   4  5 class.error
1 1 1   0   0  0   0.5000000
2 0 2   7   0  0   0.7777778
3 0 3 100  12  0   0.1304348
4 0 0  15 105  4   0.1532258
5 0 0   0   7 28   0.2000000
OOB estimate of error rate: 17.19% というのは、予測結果の誤りの割合

importance(jpn.RFmodel)
      MeanDecreaseGini
Token        52.734429
Type         38.862079
NoS          14.514875
TTR          16.444470
GI           16.044448
MATTR         9.509776
AWL          15.182801
ASL          16.387543
各要因の影響力の強さ（ジニ係数）
}}
{{ref_image Gini.png}}

!この後、テストデータで予測してみる
 jpn.RFpredict = predict(jpn.RFmodel, newdata)