RetroC temporal classification challenge
Guess the publication year of a Polish text. [ver. 1.0.0]
Log in to write a comment
kubapok
2020-01-04 23:30
submitted a solution:xgb standard
kubapok
2020-01-04 23:24
submitted a solution:lr
[anonymized]
2017-05-29 15:18
submitted a solution:Basic linear regression
[anonymized]
2017-05-28 13:55
submitted a solution:50 ep ,00001 lr, 1814-2014
[anonymized]
2017-05-19 19:18
submitted a solution:20 ep ,00005 lr
[anonymized]
2017-05-19 18:52
submitted a solution:10 ep ,0001 lr
[anonymized]
2016-06-11 14:29
submitted a solution:Max entropy solution for 4 grams and 5 class
[anonymized]
2016-06-11 14:09
submitted a solution:neural network, 4grams, 8 class
[anonymized]
2016-06-11 14:03
submitted a solution:GLM
[anonymized]
2016-06-11 13:16
submitted a solution:GLM solution for 4 grams and 8 class
[anonymized]
2016-05-29 14:08
submitted a solution:VW -nn 6 on up to 4-grams and [5-7] tokens + wiki years
[anonymized]
2016-05-29 10:35
submitted a solution:Wiki years - min year from article, with [184-2013]
[anonymized]
2016-05-29 08:21
submitted a solution:Wiki years
[anonymized]
2016-05-29 07:48
submitted a solution:Wiki years
[anonymized]
2016-05-25 13:13
submitted a solution:VW -nn 6 on up to 5-grams and [6-8] tokens
[anonymized]
2016-05-25 11:40
submitted a solution:Current best solution with changed learning rate
[anonymized]
2016-05-13 13:21
submitted a solution:Max entropy solution for 4 grams and 8 class
[anonymized]
2016-05-13 13:16
submitted a solution:Solution for 4grams and 8 class
[anonymized]
2016-05-07 21:29
submitted a solution:Solution for 4grams and 10 class
[anonymized]
2016-05-07 21:25
submitted a solution:Solution for 4grams and 20 class
[anonymized]
2016-04-30 18:20
submitted a solution:Solution for 4grams and 40 class
[anonymized]
2016-04-29 20:21
submitted a solution:Solution for 4grams and 67 class
[anonymized]
2016-04-29 07:27
submitted a solution:Solution for 4grams and 100 class
[anonymized]
2016-04-22 11:32
submitted a solution:Solution for 4grams
[anonymized]
2016-04-20 12:19
submitted a solution:Solution
[anonymized]
2016-01-07 07:56
submitted a solution:Po najczęściej występujących słowach
p/tlen
2015-12-19 11:07
submitted a solution:VW classifier (10-year chronon)
[anonymized]
2015-12-16 23:38
submitted a solution:first submission
[anonymized]
2015-12-16 20:40
submitted a solution:another try with more bits
[anonymized]
2015-12-16 20:13
submitted a solution:+ script
[anonymized]
2015-12-14 08:01
submitted a solution:tfidf, min_df = 3
[anonymized]
2015-12-13 22:16
submitted a solution:test with min_df = 3
[anonymized]
2015-12-13 18:50
submitted a solution:Simple 2-layer network with up to 4 chars n-grams
p/tlen
2015-12-13 14:31
submitted a solution:VW -nn 6 on up to 4-grams and [5-7] tokens
p/tlen
2015-12-13 12:50
submitted a solution:VW -nn 6 on up to 4-grams and [5-8] tokens
p/tlen
2015-12-13 12:49
submitted a solution:VW -nn 6 on up to 4-grams and [5-8] tokens
[anonymized]
2015-12-13 11:52
submitted a solution:Corrected better solution
[anonymized]
2015-12-13 11:43
submitted a solution:Maybe better solution
[anonymized]
2015-12-12 22:02
submitted a solution:The same as last, best epoch
[anonymized]
2015-12-12 15:56
submitted a solution:test samego dev-0 dla okrojonego treningu
p/tlen
2015-12-12 15:12
submitted a solution:up to 4-grams (VW with -nn 6 found with vw-hypersearch)
p/tlen
2015-12-12 15:10
submitted a solution:up to 4-grams (VW with -nn 6 found with vw-hypersearch)
[anonymized]
2015-12-11 19:49
submitted a solution:Same as lat, word2vec pretraining
[anonymized]
2015-12-11 19:37
submitted a solution:200 hidden units in GRU, more epochs, full batches
[anonymized]
2015-12-11 16:32
submitted a solution:First solution
p/tlen
2015-12-10 21:15
submitted a solution:up to 4-grams (VW with -nn 10)
[anonymized]
2015-12-09 08:46
submitted a solution:900 words, more units, RMSprop
p/tlen
2015-12-08 20:18
submitted a solution:signi tempori (with poor man's stemming by taking the first 7 letters)
[anonymized]
2015-12-07 21:26
submitted a solution:Better regression layer, 10 year bins weighted average
[anonymized]
2015-12-07 14:58
submitted a solution:just one epoch, to avoid overfitting
[anonymized]
2015-12-07 14:50
submitted a solution:Much simpler model, less hidden units, l2=1e-5
[anonymized]
2015-12-07 09:20
submitted a solution:now with better regularization
[anonymized]
2015-12-04 22:47
submitted a solution:Neural Network, first try
p/tlen
2015-11-25 22:29
submitted a solution:up to 4-grams
p/tlen
2015-11-25 22:27
submitted a solution:up to 4-grams
p/tlen
2015-11-25 20:47
submitted a solution:5-grams
p/tlen
2015-11-25 20:05
submitted a solution:birth years
p/tlen
2015-11-24 22:23
submitted a solution:hand-crafted rules
p/tlen
2015-11-24 21:40
submitted a solution:by known words
p/tlen
2015-11-24 20:24
submitted a solution:by year references
p/tlen
2015-11-24 19:59
submitted a solution:null model with a half year
p/tlen
2015-11-24 19:58
submitted a solution:null model
p/tlen
2015-11-14 16:30
submitted a solution:test