"He Said She Said" classification challenge
Guess whether a text in Polish was written by a man or woman. [ver. 1.0.0]
This is a long list of all submissions, if you want to see only the best, click leaderboard.
| # | submitter | when | ver. | description | dev-0 Accuracy | test-A Accuracy | |
|---|---|---|---|---|---|---|---|
| 85 | [anonymized] | 2019-02-18 18:06 | 1.0.0 | My solution | N/A | N/A | |
| 84 | [anonymized] | 2017-02-27 17:02 | 1.0.0 | moje rozwiazanie 1 | N/A | N/A | |
| 25 | [anonymized] | 2016-06-23 09:01 | 1.0.0 | 3gram model KenLM + stemming | 0.6688909559051509 | 0.6564756690423372 | |
| 5 | [anonymized] | 2016-06-19 20:27 | 1.0.0 | New 3 best voting | 0.7192677370216093 | 0.70309623825931 | |
| 7 | [anonymized] | 2016-06-18 15:05 | 1.0.0 | LSTM + ReLU + Softmax | 0.6999905636214125 | 0.6932257988389107 | |
| 6 | [anonymized] | 2016-06-18 09:25 | 1.0.0 | RNN + more layers + ReLU | 0.702632749625915 | 0.6965415113041016 | |
| 26 | [anonymized] | 2016-06-17 21:37 | 1.0.0 | KenLM + stemming + 2grams | 0.6672126285706582 | 0.6561393779204229 | |
| 4 | [anonymized] | 2016-06-17 15:47 | 1.0.0 | RNN + more layers | 0.7067982367452582 | 0.70309623825931 | |
| 3 | [anonymized] | 2016-06-17 09:52 | 1.0.0 | 3 best voting | 0.7142799369110688 | 0.7047009958937084 | |
| 24 | [anonymized] | 2016-06-17 09:05 | 1.0.0 | 3gram model test-A KenLM + stemming | 0.6672193688410779 | 0.6564756690423372 | |
| 27 | [anonymized] | 2016-06-17 09:03 | 1.0.0 | 2gram model KenLM + stemming testA | 0.6672193688410779 | 0.6560803794779818 | |
| 8 | [anonymized] | 2016-06-17 09:00 | 1.0.0 | RNN - LSTM - 2 layers - 3 epch | 0.6929402407624594 | 0.6872964553735781 | |
| 80 | [anonymized] | 2016-06-17 08:59 | 1.0.0 | 2gram model KenLM + stemming | 0.6672193688410779 | N/A | |
| 79 | [anonymized] | 2016-06-16 09:56 | 1.0.0 | KenLM + stemming + 3grams | 0.671674687588466 | N/A | |
| 16 | [anonymized] | 2016-06-13 13:36 | 1.0.0 | Both directions LSTM | 0.680457259945269 | 0.6733846226459621 | |
| 37 | [anonymized] | 2016-06-09 12:42 | 1.0.0 | 1M words + CONVOLUTION + LSTM | 0.6549790377589949 | 0.6441095955066787 | |
| 9 | [anonymized] | 2016-06-06 22:21 | 1.0.0 | LSTM + 1M words | 0.6945781264744342 | 0.6861518855902204 | |
| 51 | [anonymized] | 2016-06-04 15:54 | 1.0.0 | RNN - GRU & 5 epoch & 50k words | 0.6198217872501045 | 0.6103860858073347 | |
| 36 | p/tlen | 2016-05-30 20:51 | 1.0.0 | simple NN trained on all (3 passes) with logistic regression | 0.6506450438791604 | 0.6452659649785246 | |
| 38 | p/tlen | 2016-05-30 19:40 | 1.0.0 | simple NN trained on all (3 passes) | 0.6471738046130411 | 0.6428057299287299 | |
| 39 | p/tlen | 2016-05-30 10:41 | 1.0.0 | simple NN train on all | 0.6462503875655491 | 0.6408174824184641 | |
| 41 | p/tlen | 2016-05-30 07:48 | 1.0.0 | simple NN trained on 1M utterances | 0.6410199377199013 | 0.6348409401991787 | |
| 78 | p/tlen | 2016-05-30 07:46 | 1.0.0 | simple NN with 1M utterances | 0.6410199377199013 | N/A | |
| 77 | p/tlen | 2016-05-29 20:09 | 1.0.0 | skeleton for NN solutions | 0.4042949003114005 | 0.4082692216925473 | |
| 18 | [anonymized] | 2016-05-25 06:51 | 1.0.0 | Doc2vec + 50k words + LR | 0.6672598104635958 | 0.6597028838438665 | |
| 62 | [anonymized] | 2016-05-24 19:09 | 1.0.0 | Doc2vec + LR | 0.6038540866259554 | 0.5844621701987067 | |
| 40 | [anonymized] | 2016-05-24 17:13 | 1.0.0 | lemma + nozdi naive bayes | 0.6604386567989108 | 0.6398853070278945 | |
| 29 | [anonymized] | 2016-05-24 11:55 | 1.0.0 | klon rozwiazania Mateusza + RandomForestClassifier | 0.6727598711260295 | 0.6529239628073819 | |
| 17 | [anonymized] | 2016-05-23 19:11 | 1.0.0 | Logistic Regression + Hashing Vectorizer - in memory | 0.6800124020975722 | 0.6708181903997734 | |
| 23 | [anonymized] | 2016-05-23 18:19 | 1.0.0 | ANN + 3gram + hashing vectorizer | 0.6602364486863213 | 0.6567057629678577 | |
| 70 | [anonymized] | 2016-05-23 16:04 | 1.0.0 | Another 100k sample with RandomForestClassifier | 0.5532616168560682 | 0.5534997876056073 | |
| 49 | [anonymized] | 2016-05-23 13:20 | 1.0.0 | more iterations + randomized start | 0.6214664132324988 | 0.6135189031009581 | |
| 19 | [anonymized] | 2016-05-23 04:15 | 1.0.0 | nozdi Naive Bayes + Tfidf + swear words + emoticons | 0.6763996171526402 | 0.6583164204465002 | |
| 21 | [anonymized] | 2016-05-22 13:15 | 1.0.0 | nozdi Naive Bayes + Tfidf + swear words + emoticons v2 | 0.6763726560709615 | 0.658304620758012 | |
| 68 | [anonymized] | 2016-05-18 20:07 | 1.0.0 | 100k sample TFIDF + RFC | 0.5575821301950634 | 0.5588155472695521 | |
| 69 | [anonymized] | 2016-05-18 18:21 | 1.0.0 | 100k sample CV + RFC | 0.5593548213154311 | 0.5580898664275263 | |
| 63 | [anonymized] | 2016-05-18 14:16 | 1.0.0 | 1mln sample with RandomForestClassifier | 0.580188997182567 | 0.583683390758484 | |
| 20 | [anonymized] | 2016-05-17 05:14 | 1.0.0 | nozdi Naive Bayes + Tfidf + swear words + emoticons | 0.6763726560709615 | 0.658304620758012 | |
| 66 | [anonymized] | 2016-05-16 21:54 | 1.0.0 | 400k sample RandomTreeClassifier | 0.5717569188875857 | 0.5713822155095105 | |
| 67 | [anonymized] | 2016-05-16 19:44 | 1.0.0 | 200k sample size with RandomForestClassifier | 0.5654008438818565 | 0.5650457827913343 | |
| 47 | [anonymized] | 2016-05-16 16:08 | 1.0.0 | 100k samle with RandomForestClassifier | 0.6359377738234858 | 0.6191768537310615 | |
| 1 | p/tlen | 2016-05-15 18:31 | 1.0.0 | VW tokens + 3-gram LM (+fix for the latest grep) | 0.7231029508903897 | 0.7107778354651437 | |
| 60 | [anonymized] | 2016-05-14 14:55 | 1.0.0 | Added naive_bayes.py | 0.6201453202302476 | 0.6006985415585029 | |
| 59 | [anonymized] | 2016-05-14 14:44 | 1.0.0 | char ngrams, sample train on 100k | 0.6201453202302476 | 0.6006985415585029 | |
| 50 | [anonymized] | 2016-05-09 16:36 | 1.0.0 | Logistic regression (partial fit, iter_n=1, alpha=0.00005) | 0.6213113870128469 | 0.6129112191438146 | |
| 58 | [anonymized] | 2016-05-09 16:06 | 1.0.0 | Logistic regression (partial fit, iter_n=5) | 0.6084913926746741 | 0.6022030018407514 | |
| 56 | [anonymized] | 2016-05-08 21:35 | 1.0.0 | Logistic regression (partial fit) | 0.6088081853843976 | 0.6023091990371454 | |
| 57 | [anonymized] | 2016-05-08 21:28 | 1.0.0 | Logistic regression (partial fit) | 0.6088081853843976 | 0.6023091990371454 | |
| 22 | [anonymized] | 2016-05-07 21:24 | 1.0.0 | nozdi Naive Bayes + Tfidf | 0.6763726560709615 | 0.658304620758012 | |
| 31 | [anonymized] | 2016-05-04 19:04 | 1.0.0 | Ensemble Multinomial NB+ BernoulliNB | 0.6673676547903102 | 0.6473486099966961 | |
| 30 | [anonymized] | 2016-04-23 12:29 | 1.0.0 | Naive bayes | 0.6727666113964492 | 0.6529239628073819 | |
| 28 | [anonymized] | 2016-04-23 09:10 | 1.0.0 | Naive bayes | 0.6727666113964492 | 0.6529239628073819 | |
| 44 | [anonymized] | 2016-04-23 08:43 | 1.0.0 | Naive bayes | 0.6727666113964492 | 0.6317848208807287 | |
| 45 | [anonymized] | 2016-04-22 09:21 | 1.0.0 | Naive bayes with stop words | 0.6723824159825292 | 0.6293186859866899 | |
| 35 | [anonymized] | 2016-04-20 06:55 | 1.0.0 | Naive bayes | 0.6640514417438428 | 0.6463220370982206 | |
| 34 | [anonymized] | 2016-04-19 17:57 | 1.0.0 | Naive bayes | 0.6640514417438428 | 0.6463220370982206 | |
| 61 | p/tlen | 2016-03-24 22:01 | 1.0.0 | Vowpal Wabbit -nn 6 on morphosyntactic tags | 0.5950243323762149 | 0.5917425779959409 | |
| 55 | p/tlen | 2016-03-24 21:40 | 1.0.0 | 6-gram LM on morphosyntactic tags | 0.598711260295763 | 0.6058255062066361 | |
| 54 | p/tlen | 2016-03-24 21:21 | 1.0.0 | 6-gram LM on morphosyntactic tags | 0.6895701055526348 | 0.6058255062066361 | |
| 2 | p/tlen | 2016-02-20 12:48 | 1.0.0 | VW tokens + 3-gram LM | 0.7232040549466845 | 0.7106657384245056 | |
| 13 | p/tlen | 2016-02-19 21:30 | 1.0.0 | 3-gram language model | 0.6895701055526348 | 0.6798626516259971 | |
| 10 | p/tlen | 2016-02-19 20:28 | 1.0.0 | VW tokens + 300 V2W classes | 0.6910866663970558 | 0.6850899136262802 | |
| 11 | p/tlen | 2016-02-19 13:15 | 1.0.0 | VW tokens + 5-suffixes + NN | 0.6910057831520201 | 0.6842993344975693 | |
| 14 | p/tlen | 2016-02-19 07:51 | 1.0.0 | VW tokens + 5-prefixes + NN | 0.6844070584111834 | 0.6792903667343182 | |
| 12 | p/tlen | 2016-02-18 20:05 | 1.0.0 | Vowpal Wabbit on tokens only + small NN | 0.6883838179587765 | 0.6832727615990938 | |
| 15 | p/tlen | 2016-02-18 19:53 | 1.0.0 | Vowpal Wabbit on tokens only | 0.6793518555964465 | 0.6754908670411102 | |
| 43 | [anonymized] | 2016-02-15 18:27 | 1.0.0 | Fixed source code, added makefile. | 0.6402919885145792 | 0.6340503610704677 | |
| 53 | [anonymized] | 2016-02-15 18:16 | 1.0.0 | Fixed source code, added makefile.' | 0.6137353230611612 | 0.6092238164912447 | |
| 52 | [anonymized] | 2016-02-15 16:47 | 1.0.0 | Logistic regression on words, punctuation n-grams, and suffixes. | 0.6137353230611612 | 0.6092238164912447 | |
| 42 | [anonymized] | 2016-02-13 23:09 | 1.0.0 | Logistic regression on words and punctuation n-grams. | N/A | 0.6340503610704677 | |
| 83 | [anonymized] | 2016-02-13 23:00 | 1.0.0 | Logistic regression on words and punctuation n-grams. | N/A | N/A | |
| 48 | [anonymized] | 2016-02-12 09:24 | 1.0.0 | Logistic regression, where features are unique, lowercased words longer than one character,truncated after the 5th if necessary. | N/A | 0.6157431443809883 | |
| 73 | [anonymized] | 2016-02-12 07:55 | 1.0.0 | men only baseline | 0.5 | 0.5 | |
| 71 | p/tlen | 2016-02-11 22:10 | 1.0.0 | use "leaks" | 0.5099823404915005 | 0.5103896257138811 | |
| 72 | [anonymized] | 2016-01-08 09:25 | 1.0.0 | man only baseline | 0.5 | 0.5 | |
| 82 | [anonymized] | 2015-12-17 08:38 | 1.0.0 | naive bayes | 0.577344603065475 | N/A | |
| 46 | [anonymized] | 2015-12-17 07:34 | 1.0.0 | pliki zrodlowe w odp. folderach | 0.6269058114611559 | 0.6288407986029169 | |
| 65 | [anonymized] | 2015-12-16 21:17 | 1.0.0 | Dodane kody zrodlowe | 0.5852779013494022 | 0.5719132014914806 | |
| 64 | [anonymized] | 2015-12-16 21:09 | 1.0.0 | uhhh, test numer 3 | 0.5852779013494022 | 0.5719132014914806 | |
| 75 | [anonymized] | 2015-12-16 18:56 | 1.0.0 | dominik szczeszynski - proba 2 | 0.5009908197516885 | 0.4991504224288479 | |
| 74 | [anonymized] | 2015-12-10 09:08 | 1.0.0 | Piotr Mizerka kobieta_czy_mezczyzna Naiwny Bayes | 0.6269058114611559 | 0.4995044130834946 | |
| 33 | [anonymized] | 2015-12-10 09:05 | 1.0.0 | naive bayes by Przemysław Nowaczyk kod zrodlowy i zasoby | 0.6662487699006484 | 0.646864822768679 | |
| 81 | [anonymized] | 2015-12-10 09:05 | 1.0.0 | Piotr Mizerka kobieta_czy_mezczyzna Naiwny Bayes | 0.6269058114611559 | N/A | |
| 32 | [anonymized] | 2015-12-10 08:19 | 1.0.0 | naive bayes by Przemysław Nowaczyk | 0.6662487699006484 | 0.646864822768679 | |
| 76 | [anonymized] | 2015-12-10 07:48 | 1.0.0 | dominik szczeszynski - solution | 0.4977824510319354 | 0.49843654127531034 |