Table 4 Top 25 word stems for BOWs according to the variable importance extracted from scikit-learn’s ExtraTreesClassifier and stemmed using nltk’s SnowballStemmer [17, 22]
From: Word2Vec inversion and traditional text classifiers for phenotyping lupus
Rank | Word | VIMP |
|---|---|---|
1 | C3 | 0.0311 |
2 | Sle | 0.0225 |
3 | Graviti | 0.0172 |
4 | Sole | 0.0126 |
5 | Phurin | 0.0113 |
6 | Epitheli | 0.0084 |
7 | C4 | 0.0065 |
8 | Yet | 0.0065 |
9 | Lymph | 0.0063 |
10 | Hemlymph | 0.0059 |
11 | Educ | 0.0059 |
12 | Resolv | 0.0054 |
13 | 912 | 0.0054 |
14 | Fatigu | 0.0050 |
15 | Thrombocytopenia | 0.0047 |
16 | 2500 | 0.0047 |
17 | Need | 0.0047 |
18 | Naugl | 0.0047 |
19 | Clot | 0.0043 |
20 | Screen | 0.0042 |
21 | Antidoubl | 0.0040 |
22 | Beat | 0.0040 |
23 | Acut | 0.0038 |
24 | 843identificationremov | 0.0038 |
25 | Pregnanc | 0.0036 |