RetroC2 temporal classification challenge

Guess the publication year of a Polish text. [ver. 1.0.0]

# submitter when ver. description dev-0 RMSE dev-1 RMSE test-A RMSE
15 kubapok 2020-06-30 22:45 1.0.0 fasttext as classification problem Mikolaj Bachorz experiment reproduction 50 epochs devs included 11.250 10.806 20.805
22 kubapok 2020-06-30 21:19 1.0.0 fasttext as classification problem Mikolaj Bachorz experiment reproduction 50 epochs 21.670 19.827 24.101
35 kubapok 2020-06-30 13:35 1.0.0 fasttext as classification problem Mikolaj Bachorz experiment reproduction 33.336 25.245 30.931
103 Anna Maduzia 2020-06-24 00:12 1.0.0 xgboost solution ready-made xgboost 54.560 52.444 54.317
123 Anna Maduzia 2020-06-18 11:49 1.0.0 xgboost skrypt + pliki out ready-made xgboost 66.925 64.530 67.634
1 kubapok 2020-06-02 17:12 1.0.0 linear layer on top of polish roberta- Adam lr 1e-07 12.260 12.317 13.328
2 kubapok 2020-05-28 15:09 1.0.0 linear layer on top of polish roberta (both finetuned 2 epochs) 13.671 13.723 15.606
14 kubapok 2020-05-24 18:27 1.0.0 xgb on top of polish roberta mean token 17.685 17.286 20.458
8 Mikolaj Bachorz 2020-05-18 08:34 1.0.0 v3 0.913 1.048 18.923
5 Mikolaj Bachorz 2020-05-18 05:20 1.0.0 v2 fasttext 1.084 1.216 17.989
3 kubapok 2020-01-28 19:54 1.0.0 ensemble of 4, bilstm 15.326 15.060 17.262
7 kubapok 2020-01-28 19:52 1.0.0 ensemble of 4, bilstm 16.342 16.339 18.758
6 kubapok 2020-01-19 22:49 1.0.0 keras bilstm lstm 16.342 16.339 18.758
21 kubapok 2020-01-12 21:38 1.0.0 keras fasttext like 22.069 20.969 23.782
17 kubapok 2020-01-09 21:38 1.0.0 tfidf baseline maxminclipping linear-regression tf-idf 21.327 21.222 23.114
18 kubapok 2020-01-09 16:16 1.0.0 tfidf lr baseline tf-idf 21.673 21.317 23.167
112 [anonymised] 2019-07-21 14:25 1.0.0 CNN 30,31,32 72.250 52.565 57.806
116 [anonymised] 2019-07-16 18:50 1.0.0 CNN, lr 0.001, 100 filters, [4,5,6,7] filter sizes, dropout 0.5 76.269 N/A 59.799
115 [anonymised] 2019-07-16 18:17 1.0.0 CNN, lr 0.001, 100 filters, [4,5,6,7] filter sizes, dropout 0.5 N/A N/A 59.799
90 [anonymised] 2019-07-13 20:29 1.0.0 Feedforward, word embeddings NKJP + Wikipedia, model01 61.224 43.289 47.333
89 [anonymised] 2019-07-13 20:15 1.0.0 Feedforward, word embeddings NKJP + Wikipedia, model01 61.224 N/A 47.333
88 [anonymised] 2019-07-13 19:20 1.0.0 Feedforward, word embeddings NKJP + Wikipedia, model48 N/A N/A 47.333
82 [anonymised] 2019-07-13 19:06 1.0.0 Feedforwar, word embeddings NKJP + Wikipedia N/A N/A 47.045
128 [anonymised] 2019-07-13 12:19 1.0.0 Char CNN 30e, 0.001lr 69.770 45.606 74.298
110 [anonymised] 2019-06-12 14:49 1.0.0 wordlist 4 63.600 55.405 56.295
93 [anonymised] 2019-06-12 14:40 1.0.0 wordlist 4 57.742 51.007 51.853
101 [anonymised] 2019-06-12 14:16 1.0.0 wordlist 4 64.766 50.656 53.026
104 [anonymised] 2019-06-12 14:12 1.0.0 wordlist 4 63.817 53.179 54.668
102 [anonymised] 2019-06-12 13:56 1.0.0 wordlist 4 64.448 51.292 53.421
119 [anonymised] 2019-06-12 12:20 1.0.0 wordlist 4 65.050 63.803 63.186
96 [anonymised] 2019-06-12 12:00 1.0.0 wordlist 4 65.818 49.338 52.340
94 [anonymised] 2019-06-12 11:27 1.0.0 wordlist 4 66.698 48.749 52.150
130 [anonymised] 2019-06-12 11:13 1.0.0 wordlist + random choice 2 83.138 78.207 79.643
55 [anonymised] 2019-06-10 15:23 1.0.0 graf self-made linear-regression graph 48.238 41.755 39.246
87 [anonymised] 2019-06-08 11:23 1.0.0 Bayes to predict some time range fix naive-bayes 41.561 50.584 47.205
27 [anonymised] 2019-06-04 15:41 1.0.0 Vowpal Wabbit quadratic model + graph v2 vowpal-wabbit graph 27.479 23.632 27.297
33 [anonymised] 2019-06-04 15:05 1.0.0 Vowpal Wabbit quadratic model + graph vowpal-wabbit graph 28.930 25.384 28.401
28 [anonymised] 2019-06-04 12:03 1.0.0 vw first encounter(loss function, -b 27, passes=20, quadratic model) vowpal-wabbit graph 26.939 23.654 27.384
23 [anonymised] 2019-06-01 15:50 1.0.0 tf ready-made linear-regression tf 23.149 21.791 24.291
81 [anonymised] 2019-05-29 06:10 1.0.0 Test feedforward 69.770 45.606 46.987
127 [anonymised] 2019-05-27 14:04 1.0.0 Bayes to predict some time range naive-bayes 65.137 82.283 71.060
129 [anonymised] 2019-05-27 13:06 1.0.0 naive bayes naive-bayes 69.913 89.059 77.466
92 [anonymised] 2019-05-22 08:38 1.0.0 BiLSTM 30epochs 22nd new tokenizer 69.770 45.606 50.381
140 [anonymised] 2019-05-22 08:36 1.0.0 BiLSTM 3- ep0 epochs 22nd new tokenizer 69.770 45.606 N/A
117 [anonymised] 2019-05-21 19:00 1.0.0 BiLSTM 3- epochs 22nd new tokenizer 69.770 45.606 62.489
30 [anonymised] 2019-05-17 18:45 1.0.0 Vowpal Wabbit - linear regression + graph vowpal-wabbit graph 29.092 24.445 27.808
26 [anonymised] 2019-05-11 22:40 1.0.0 BiLSTM, 30epochs, model 28th 69.770 45.606 24.605
39 [anonymised] 2019-05-09 23:19 1.0.0 ready-made tf-df: a fix ready-made linear-regression tf-idf 29.817 26.292 33.388
25 [anonymised] 2019-05-09 17:51 1.0.0 BiLSTM, 30epochs, model 28th N/A 45.606 24.605
24 [anonymised] 2019-05-09 05:03 1.0.0 BiLSTM, 30epochs, model 28th N/A N/A 24.605
31 [anonymised] 2019-05-06 15:08 1.0.0 tfidf 3k words low range 59.446 N/A 28.277
29 [anonymised] 2019-05-06 15:05 1.0.0 tfidf 50k words low reduction range ready-made linear-regression tf-idf 59.446 N/A 27.708
70 [anonymised] 2019-05-06 15:01 1.0.0 My solution go.php rule-based 57.517 45.142 42.687
65 [anonymised] 2019-05-06 14:49 1.0.0 transfer files to VM 59.446 N/A 40.674
64 [anonymised] 2019-05-06 14:08 1.0.0 all documents to predict on vector 30k words TFIDF KNN ready-made knn tf-idf 59.446 N/A 40.674
40 [anonymised] 2019-05-06 13:57 1.0.0 all documents to predict on vector 10k words TFIDF Linear ready-made linear-regression tf-idf 59.446 N/A 34.444
77 [anonymised] 2019-05-06 13:53 1.0.0 300 documents to predict on vector 10k words TFIDF Linear self-made linear-regression tf-idf 59.446 N/A 43.571
20 [anonymised] 2019-05-03 19:49 1.0.0 BiLSTM w/o sorting N/A 43.734 23.225
36 [anonymised] 2019-05-03 18:37 1.0.0 3000 words tf-idf self-made linear-regression tf-idf 34.237 30.865 32.631
37 [anonymised] 2019-05-03 17:25 1.0.0 2500 words tf-idf 34.857 31.521 32.912
38 [anonymised] 2019-05-03 16:45 1.0.0 2000 words tf-idf self-made linear-regression tf-idf 35.672 32.114 33.384
44 [anonymised] 2019-05-03 16:27 1.0.0 1000 words tf-idf 38.925 35.224 36.004
19 [anonymised] 2019-05-03 07:17 1.0.0 BiLSTM w\o sorting N/A N/A 23.225
91 [anonymised] 2019-04-29 14:57 1.0.0 self made TFIDF 1000 documents 500 word vector KNN4 self-made linear-regression knn tf-idf 59.446 N/A 48.323
53 [anonymised] 2019-04-28 20:08 1.0.0 linner ready tf ready-made linear-regression tf 51.562 41.400 39.240
86 [anonymised] 2019-04-27 18:41 1.0.0 change encoding 59.446 N/A 47.094
85 [anonymised] 2019-04-27 17:54 1.0.0 tfidf - not ready 59.446 N/A 47.094
4 Artur Nowakowski 2019-04-19 07:52 1.0.0 optimized word2vec + nn neural-network word2vec 15.913 14.587 17.789
13 Artur Nowakowski 2019-04-17 16:02 1.0.0 wordvec + nn 5-fold validation 17.644 16.803 20.232
12 Artur Nowakowski 2019-04-17 09:20 1.0.0 word2vec + nn optimized for dev1 18.714 15.851 19.915
111 [anonymised] 2019-04-16 22:30 1.0.0 Now with CHARTS self-made linear-regression graph 62.496 57.478 57.119
43 [anonymised] 2019-04-16 16:36 1.0.0 simple lin reg self-made linear-regression graph N/A N/A 35.958
52 [anonymised] 2019-04-16 16:24 1.0.0 Figure add self-made linear-regression graph N/A 40.479 38.431
10 Artur Nowakowski 2019-04-16 16:13 1.0.0 basic word2vec + nn solution (optimized for dev0) 16.793 17.702 19.557
51 [anonymised] 2019-04-16 15:36 1.0.0 years in text self-made linear-regression graph 50.710 N/A 38.419
134 [anonymised] 2019-04-16 15:20 1.0.0 mean year found in text 1341469.078 N/A 1345919.117
57 [anonymised] 2019-04-16 13:24 1.0.0 Regresja liniowa (USA |usa |stany zjednoczone) + Lata 48.599 42.924 39.382
42 [anonymised] 2019-04-16 13:14 1.0.0 simple linear regression self-made linear-regression graph N/A N/A 35.958
56 [anonymised] 2019-04-16 13:10 1.0.0 Regresja liniowa (USA|usa|stany zjednoczone) plus lata N/A 42.924 39.382
67 [anonymised] 2019-04-16 09:49 1.0.0 linear regression self-made linear-regression graph 61.474 40.228 41.220
84 [anonymised] 2019-04-16 01:12 1.0.0 Merge branch 'master' of ssh://gonito.net/huntekah/retroc2 self-made linear-regression graph 59.446 N/A 47.094
83 [anonymised] 2019-04-16 00:56 1.0.0 Honest one variable solution, without fancy, and thus easy methods self-made linear-regression graph 59.446 N/A 47.094
32 [anonymised] 2019-04-15 20:33 1.0.0 Basic ready-made solution with one column ready-made linear-regression tf-idf 30.212 26.677 28.298
72 [anonymised] 2019-04-15 18:40 1.0.0 excel plots :) self-made linear-regression graph 52.092 46.796 42.991
62 [anonymised] 2019-04-15 18:39 1.0.0 XXDDDD my solution self-made linear-regresion ADD CHARTS selfMadeLinearRegres_Solver.py self-made linear-regression graph 48.699 42.928 39.904
63 [anonymised] 2019-04-15 16:35 1.0.0 one variable regression self-made linear-regression graph N/A N/A 40.219
71 [anonymised] 2019-04-15 12:46 1.0.0 hope that is final one self-made linear-regression 52.092 46.796 42.991
139 [anonymised] 2019-04-15 12:40 1.0.0 now more iterations 52.092 46.796 N/A
132 [anonymised] 2019-04-15 12:02 1.0.0 forgot to add out files xd 279.733 285.868 269.460
133 [anonymised] 2019-04-15 11:57 1.0.0 date detection and linear regression 271.215 296.533 271.270
61 [anonymised] 2019-04-15 11:54 1.0.0 XDDD my solution self-made linear-regresion selfMadeLinearRegres_Solver.py self-made linear-regression 48.699 42.928 39.904
16 Artur Nowakowski 2019-04-15 11:40 1.0.0 Linear regression with TF-IDF weighing scheme ready-made linear-regression tf-idf 20.750 20.676 22.341
34 [anonymised] 2019-04-11 18:56 1.0.0 Basic ready-made solution ready-made linear-regression tf-idf 30.723 26.963 28.668
54 [anonymised] 2019-04-10 10:28 1.0.0 wykrywanie dat top prio rule-based 48.238 41.755 39.246
138 [anonymised] 2019-04-09 19:11 1.0.0 not many rules rule-based N/A N/A N/A
107 [anonymised] 2019-04-09 16:55 1.0.0 post OCR signs 69.841 N/A 55.400
106 [anonymised] 2019-04-09 16:53 1.0.0 post OCR signs 73.236 N/A 55.400
126 [anonymised] 2019-04-09 16:51 1.0.0 post OCR signs 73.236 N/A 69.970
125 [anonymised] 2019-04-09 16:46 1.0.0 post OCR signs 351.237 N/A 69.970
109 [anonymised] 2019-04-09 16:43 1.0.0 solution with simple word list2 rule-based 63.523 55.172 56.124
122 [anonymised] 2019-04-09 16:28 1.0.0 post OCR signs N/A N/A 66.046
69 [anonymised] 2019-04-09 16:15 1.0.0 My solution basic rule-based solver.py rule-based 57.517 45.142 42.682
68 [anonymised] 2019-04-09 16:12 1.0.0 fourth solution rule-based N/A 43.947 41.235
47 [anonymised] 2019-04-09 15:28 1.0.0 bad solution 2 rule-based N/A N/A 37.197
114 [anonymised] 2019-04-09 13:07 1.0.0 Bad rule-based solution rule-based N/A N/A 59.748
50 [anonymised] 2019-04-08 21:02 1.0.0 better stupid solution rule-based 50.229 39.773 38.368
49 [anonymised] 2019-04-08 19:59 1.0.0 stupid solution rule-based 50.234 41.683 38.177
66 [anonymised] 2019-04-08 18:10 1.0.0 my very simple solution3 rule-based 50.607 44.078 40.838
59 [anonymised] 2019-04-08 15:24 1.0.0 rulebased rule-based N/A N/A 39.781
75 [anonymised] 2019-04-08 15:23 1.0.0 all rules rule-based 54.064 N/A 43.101
74 [anonymised] 2019-04-08 15:21 1.0.0 slowa 57.733 N/A 43.101
58 [anonymised] 2019-04-08 12:54 1.0.0 Based on a list with years rule-based 48.666 43.031 39.695
137 [anonymised] 2019-04-08 12:42 1.0.0 Improve guessing accuracy for de0 dev1 48.666 43.031 N/A
108 [anonymised] 2019-04-08 11:32 1.0.0 my very simple solution2A 63.411 49.901 55.583
136 [anonymised] 2019-04-07 22:13 1.0.0 my very simple solution1 63.411 49.901 N/A
73 [anonymised] 2019-04-07 19:51 1.0.0 complicated rules make Good/Bad results 57.733 N/A 43.101
113 [anonymised] 2019-04-07 18:17 1.0.0 most popular words in 10-year periods java rule-based 51.117 54.080 57.981
95 [anonymised] 2019-04-07 09:44 1.0.0 simple rule based rule-based 57.733 N/A 52.306
121 [anonymised] 2019-04-07 08:42 1.0.0 my best solution rule-based 86.441 61.280 65.953
76 [anonymised] 2019-04-06 16:59 1.0.0 based on historical word list rule-based 57.220 46.287 43.424
60 [anonymised] 2019-04-05 19:22 1.0.0 simple set solution rule-based 53.933 37.245 39.885
46 Artur Nowakowski 2019-04-03 11:52 1.0.0 simple solution rule-based 50.255 37.596 36.563
105 [anonymised] 2019-04-02 16:50 1.0.0 LSTM EPOCHS=10 LR=0.001 DROPOUT=0.1 - input w/o filtering, adding missing line neural-network bilstm 66.348 N/A 55.115
120 [anonymised] 2019-03-30 19:01 1.0.0 LSTM EPOCHS=10 LR=0.001 DROPOUT=0.2 - input w/o filtering, adding missing line neural-network bilstm 73.227 N/A 63.913
131 [anonymised] 2019-03-27 17:36 1.0.0 LSTM EPOCHS=10 LR=0.001 DROPOUT=0.5 - input w/o filtering, adding missing line neural-network bilstm N/A N/A 97.902
100 [anonymised] 2019-02-19 09:08 1.0.0 Modyfikacja skryptu do uruchomienia 72.529 N/A 52.950
99 [anonymised] 2019-02-19 08:50 1.0.0 5 epochs; filtered input; feedforward network neural-network 72.529 N/A 52.950
118 p/tlen 2018-08-30 20:16 1.0.0 tescik 5 stupid N/A N/A 62.502
124 p/tlen 2018-08-30 19:38 1.0.0 tescik 3 stupid N/A N/A 69.597
79 p/tlen 2018-08-30 19:31 1.0.0 tescik2 stupid N/A N/A 46.513
80 p/tlen 2018-08-30 19:27 1.0.0 test stupid N/A N/A 46.591
98 [anonymised] 2018-08-14 18:29 1.0.0 dev0 first solution stupid neural-network 57.863 N/A 52.721
45 p/tlen 2018-05-29 07:19 1.0.0 try xgboost xgboost 39.907 37.228 36.484
11 p/tlen 2017-07-08 12:52 1.0.0 VW with yearly resolution 23.065 17.178 19.695
135 p/tlen 2017-07-08 12:50 1.0.0 VW with yearly resolution N/A N/A N/A
41 p/tlen 2017-07-07 10:23 1.0.0 year references combined with hand-crafted rules 42.302 42.654 35.807
78 p/tlen 2017-07-07 09:41 1.0.0 hand-crafted rules 48.602 50.646 44.167
48 p/tlen 2017-07-07 09:11 1.0.0 year references 46.447 39.809 37.652
9 p/tlen 2017-05-31 04:49 1.0.0 VW -nn 6 on up to 4-grams and [5-7] tokens vowpal-wabbit neural-network 22.413 16.991 19.501
97 p/tlen 2017-05-26 21:48 1.0.0 null model stupid null-model 57.735 51.906 52.539