This repository has been archived by the owner on Oct 15, 2021. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 88
/
results_old.txt
71 lines (50 loc) · 2.54 KB
/
results_old.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
CountVectorizer
LinearSVC L1 C=0.25 0.83
LinearSVC L2 C=0.062 0.838
TFIDF
LinearSVC L1 C: 1.0 0.838
LinearSVC L2 C: 1.0 0.84064462306075627,
SVC ({'C': 100.0, 'gamma': 1.0}, 0.83075610310461856
char 2-gram + word 2-gram : 83.9 char 2-gram doesn't help?
char 4-gram + word 2-gram : 0.84216553560828478
with char 1-gram ? 0.84190434936067127
char 6-gram 0.84444273315043883 {'penalty': 'l2', 'C': 0.125}
word 2-gram + features 0.84722978935040338 {'penalty': 'l2', 'C': 2.0}
features+ word 2-gram + char 6-gram 0.847486162584 {'penalty': 'l2', 'C': 0.125}
binary=true: 0.849511158172 {'penalty': 'l2', 'C': 0.25}
corrected badword list 0.852294043093
{'vect__char_max_n': 4, 'vect__word_max_n': 3, 'logr__C': 0.125, 'logr__penalty': 'l2'}
0.851030145513
{'vect__char_max_n': 4, 'vect__word_max_n': 1, 'logr__C': 0.03125, 'logr__penalty': 'l2', 'vect__char_min_n': 3}
0.852803260015
FROM NOW ON AUC!!!
0.884426825818
{'vect__char_max_n': 4, 'logr__C': 0.125, 'vect__word_max_n': 3, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l2'}
L1:
0.883239800358
{'vect__char_max_n': 4, 'logr__C': 0.25, 'vect__word_max_n': 1, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l1'}
L1 no chars:
0.881108382716
{'vect__char_max_n': 4, 'logr__C': 0.5, 'vect__word_max_n': 3, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l1'
L2 no chars:
0.883419143416
{'vect__char_max_n': 4, 'logr__C': 1.0, 'vect__word_max_n': 2, 'vect__char_min_n': 3, 'logr__class_weight': 'auto', 'logr__penalty': 'l2'}
No chars?!
0.884426825818
{'vect__char_max_n': 4, 'logr__C': 0.125, 'vect__word_max_n': 3, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l2'}
L1, no chars, !, @
0.883628749026
{'vect__char_max_n': 4, 'logr__C': 0.5, 'vect__word_max_n': 3, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l1'}
L2, no chars, !, @
0.885277997947
{'vect__char_max_n': 4, 'logr__C': 1.0, 'vect__word_max_n': 2, 'vect__char_min_n': 3, 'logr__class_weight': 'auto', 'logr__penalty': 'l2'}
L2, nochars, fixed unicode
0.887569120472
{'vect__char_max_n': 4, 'logr__C': 1.0, 'vect__word_max_n': 2, 'vect__char_min_n': 3, 'logr__class_weight': 'auto', 'logr__penalty': 'l2'}
test_prediction_12_14_41.csv
L2 with chars
{'vect__char_max_n': 4, 'logr__C': 0.0625, 'vect__word_max_n': 3, 'vect__char_min_n': 3, 'logr__class_weight': None, 'logr__penalty': 'l2'}
L1 feature selection + RF
0.89956105124940811
clf = LogisticRegression(tol=1e-8, C=0.5, penalty='l1')
{'max_features': 'log2', 'max_depth': 38, 'min_samples_leaf': 1}