RetroC temporal classification challenge

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

# submitter when ver. description dev-0 RMSE test-A RMSE
27 kubapok 2020-01-04 23:30 1.0.0 xgb standard 33.4 39.9
24 kubapok 2020-01-04 23:24 1.0.0 lr 24.6 36.9
56 [anonymized] 2017-05-29 15:18 1.0.0 Basic linear regression haskell self-made linear-regression 85.0 85.4
57 [anonymized] 2017-05-28 13:55 1.0.0 50 ep ,00001 lr, 1814-2014 self-made linear-regression N/A 93.3
59 [anonymized] 2017-05-19 19:18 1.0.0 20 ep ,00005 lr N/A 396.9
58 [anonymized] 2017-05-19 18:52 1.0.0 10 ep ,0001 lr self-made linear-regression N/A 393.3
32 [anonymized] 2016-06-11 14:29 1.0.0 Max entropy solution for 4 grams and 5 class 31.5 41.0
30 [anonymized] 2016-06-11 14:09 1.0.0 neural network, 4grams, 8 class 31.5 41.0
31 [anonymized] 2016-06-11 14:03 1.0.0 GLM 31.5 41.0
29 [anonymized] 2016-06-11 13:16 1.0.0 GLM solution for 4 grams and 8 class 31.5 41.0
2 [anonymized] 2016-05-29 14:08 1.0.0 VW -nn 6 on up to 4-grams and [5-7] tokens + wiki years 17.1 24.9
55 [anonymized] 2016-05-29 10:35 1.0.0 Wiki years - min year from article, with [184-2013] 81.8 82.9
60 [anonymized] 2016-05-29 08:21 1.0.0 Wiki years 103.1 N/A
63 [anonymized] 2016-05-29 07:48 1.0.0 Wiki years 103.1 N/A
9 [anonymized] 2016-05-25 13:13 1.0.0 VW -nn 6 on up to 5-grams and [6-8] tokens 17.9 26.4
7 [anonymized] 2016-05-25 11:40 1.0.0 Current best solution with changed learning rate 17.3 25.5
22 [anonymized] 2016-05-13 13:21 1.0.0 Max entropy solution for 4 grams and 8 class 26.6 35.8
23 [anonymized] 2016-05-13 13:16 1.0.0 Solution for 4grams and 8 class 24.7 36.8
25 [anonymized] 2016-05-07 21:29 1.0.0 Solution for 4grams and 10 class 24.7 37.4
26 [anonymized] 2016-05-07 21:25 1.0.0 Solution for 4grams and 20 class 25.1 38.4
28 [anonymized] 2016-04-30 18:20 1.0.0 Solution for 4grams and 40 class 26.7 40.3
33 [anonymized] 2016-04-29 20:21 1.0.0 Solution for 4grams and 67 class 27.3 41.6
34 [anonymized] 2016-04-29 07:27 1.0.0 Solution for 4grams and 100 class 28.4 41.8
35 [anonymized] 2016-04-22 11:32 1.0.0 Solution for 4grams 30.2 43.2
39 [anonymized] 2016-04-20 12:19 1.0.0 Solution 30.2 44.6
62 [anonymized] 2016-01-07 07:56 1.0.0 Po najczęściej występujących słowach N/A N/A
21 p/tlen 2015-12-19 11:07 1.0.0 VW classifier (10-year chronon) 24.5 34.4
54 [anonymized] 2015-12-16 23:38 1.0.0 first submission N/A 75.0
10 [anonymized] 2015-12-16 20:40 1.0.0 another try with more bits 16.7 27.2
46 [anonymized] 2015-12-16 20:13 1.0.0 + script 34.7 50.3
45 [anonymized] 2015-12-14 08:01 1.0.0 tfidf, min_df = 3 34.7 50.3
53 [anonymized] 2015-12-13 22:16 1.0.0 test with min_df = 3 41.5 60.1
13 [anonymized] 2015-12-13 18:50 1.0.0 Simple 2-layer network with up to 4 chars n-grams 16.8 27.6
1 p/tlen 2015-12-13 14:31 1.0.0 VW -nn 6 on up to 4-grams and [5-7] tokens vowpal-wabbit neural-network 17.2 24.8
5 p/tlen 2015-12-13 12:50 1.0.0 VW -nn 6 on up to 4-grams and [5-8] tokens 17.2 25.3
42 p/tlen 2015-12-13 12:49 1.0.0 VW -nn 6 on up to 4-grams and [5-8] tokens 45.9 46.5
37 [anonymized] 2015-12-13 11:52 1.0.0 Corrected better solution 31.9 44.4
36 [anonymized] 2015-12-13 11:43 1.0.0 Maybe better solution N/A 44.4
3 [anonymized] 2015-12-12 22:02 1.0.0 The same as last, best epoch 16.3 24.9
61 [anonymized] 2015-12-12 15:56 1.0.0 test samego dev-0 dla okrojonego treningu 61.5 N/A
6 p/tlen 2015-12-12 15:12 1.0.0 up to 4-grams (VW with -nn 6 found with vw-hypersearch) 17.2 25.5
41 p/tlen 2015-12-12 15:10 1.0.0 up to 4-grams (VW with -nn 6 found with vw-hypersearch) 45.9 46.5
4 [anonymized] 2015-12-11 19:49 1.0.0 Same as lat, word2vec pretraining 16.4 25.3
11 [anonymized] 2015-12-11 19:37 1.0.0 200 hidden units in GRU, more epochs, full batches 17.5 27.3
38 [anonymized] 2015-12-11 16:32 1.0.0 First solution 30.2 44.6
8 p/tlen 2015-12-10 21:15 1.0.0 up to 4-grams (VW with -nn 10) 17.7 26.0
12 [anonymized] 2015-12-09 08:46 1.0.0 900 words, more units, RMSprop 18.5 27.5
44 p/tlen 2015-12-08 20:18 1.0.0 signi tempori (with poor man's stemming by taking the first 7 letters) 48.1 50.2
15 [anonymized] 2015-12-07 21:26 1.0.0 Better regression layer, 10 year bins weighted average 19.6 30.0
20 [anonymized] 2015-12-07 14:58 1.0.0 just one epoch, to avoid overfitting 23.3 33.8
18 [anonymized] 2015-12-07 14:50 1.0.0 Much simpler model, less hidden units, l2=1e-5 20.8 31.7
17 [anonymized] 2015-12-07 09:20 1.0.0 now with better regularization 20.9 31.6
16 [anonymized] 2015-12-04 22:47 1.0.0 Neural Network, first try 22.0 30.2
14 p/tlen 2015-11-25 22:29 1.0.0 up to 4-grams 19.6 28.4
40 p/tlen 2015-11-25 22:27 1.0.0 up to 4-grams 45.9 46.5
19 p/tlen 2015-11-25 20:47 1.0.0 5-grams 22.0 33.5
49 p/tlen 2015-11-25 20:05 1.0.0 birth years 57.4 56.7
47 p/tlen 2015-11-24 22:23 1.0.0 hand-crafted rules 46.6 50.9
48 p/tlen 2015-11-24 21:40 1.0.0 by known words 57.9 56.5
43 p/tlen 2015-11-24 20:24 1.0.0 by year references 45.9 46.5
50 p/tlen 2015-11-24 19:59 1.0.0 null model with a half year 57.9 57.7
51 p/tlen 2015-11-24 19:58 1.0.0 null model 57.9 57.7
52 p/tlen 2015-11-14 16:30 1.0.0 test 81.1 57.7