{ "id": 209641, "title": "A compilation of solutions", "url": "/competitions/riiid-test-answer-prediction/discussion/209641", "author_name": "", "post_date": "2021-01-08T05:38:55.347399Z", "votes": 35, "comment_count": 8, "views": 0, "content": "

Hey all,

\n

I have made a compilation of the solutions shared so far. I will continue to update as more are released!

\n

2nd [Transformer] - https://www.kaggle.com/mamasinkgs/public-private-2nd-place-solution
\n3rd [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209585
\n6th [Custom GRU NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209581
\n11th [LSTM NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209694
\n12th [Stacked Model, LGBM, SAINT, AKT, SSAKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209635
\n13th [Stacked Transformer, LSTM, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209622
\n16th [Transformer, SAKT, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798
\n17th [Stacked Transfomer, LSTM, GRU] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209713
\n18th [Ensemble Catboost, Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209597
\n20th [SERT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209587
\n22nd [Encoder Modules] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209711
\n24th [SAINT++] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209625
\n37th [Ensemble SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209624
\n38th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209689
\n42nd [Ensemble LGB, RAINT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209596
\n52nd [Ensemble LGBM, Catboost]- https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209576
\n58th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209793
\n60th [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209586
\n62nd [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209629
\n68th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209615
\n70th [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209584
\n71st [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209577
\n76th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209599
\n125th [Ensemble LGBM, AKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209628
\n158th [LGBM, FTRL] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209619

\n

One of the things that I find interesting that I didn't expect to see is the ensembles / stacking solutions. Sorry if I missed any solutions. I will go through again once more are released and try to consolidate the ones I missed.

\n

UPDATE: I have added more model solutions and added tags in the beginning for a quick look up :) If anyone see their solution missing or misinterpreted please leave a comment and I will straighten it out.

", "messages": [ { "id": "1143871", "postDate": "01/08/2021 05:38:55", "content": "

Hey all,

\n

I have made a compilation of the solutions shared so far. I will continue to update as more are released!

\n

2nd [Transformer] - https://www.kaggle.com/mamasinkgs/public-private-2nd-place-solution
\n3rd [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209585
\n6th [Custom GRU NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209581
\n11th [LSTM NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209694
\n12th [Stacked Model, LGBM, SAINT, AKT, SSAKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209635
\n13th [Stacked Transformer, LSTM, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209622
\n16th [Transformer, SAKT, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798
\n17th [Stacked Transfomer, LSTM, GRU] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209713
\n18th [Ensemble Catboost, Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209597
\n20th [SERT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209587
\n22nd [Encoder Modules] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209711
\n24th [SAINT++] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209625
\n37th [Ensemble SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209624
\n38th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209689
\n42nd [Ensemble LGB, RAINT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209596
\n52nd [Ensemble LGBM, Catboost]- https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209576
\n58th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209793
\n60th [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209586
\n62nd [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209629
\n68th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209615
\n70th [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209584
\n71st [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209577
\n76th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209599
\n125th [Ensemble LGBM, AKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209628
\n158th [LGBM, FTRL] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209619

\n

One of the things that I find interesting that I didn't expect to see is the ensembles / stacking solutions. Sorry if I missed any solutions. I will go through again once more are released and try to consolidate the ones I missed.

\n

UPDATE: I have added more model solutions and added tags in the beginning for a quick look up :) If anyone see their solution missing or misinterpreted please leave a comment and I will straighten it out.

", "rawMarkdown": "Hey all,\n\nI have made a compilation of the solutions shared so far. I will continue to update as more are released!\n\n2nd [Transformer] - https://www.kaggle.com/mamasinkgs/public-private-2nd-place-solution\n3rd [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209585\n6th [Custom GRU NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209581\n11th [LSTM NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209694 \n12th [Stacked Model, LGBM, SAINT, AKT, SSAKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209635\n13th [Stacked Transformer, LSTM, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209622\n16th [Transformer, SAKT, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798\n17th [Stacked Transfomer, LSTM, GRU] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209713\n18th [Ensemble Catboost, Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209597\n20th [SERT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209587\n22nd [Encoder Modules] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209711\n24th [SAINT++] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209625\n37th [Ensemble SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209624\n38th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209689\n42nd [Ensemble LGB, RAINT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209596\n52nd [Ensemble LGBM, Catboost]- https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209576\n58th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209793\n60th [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209586\n62nd [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209629\n68th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209615\n70th [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209584\n71st [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209577\n76th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209599\n125th [Ensemble LGBM, AKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209628\n158th [LGBM, FTRL] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209619\n\nOne of the things that I find interesting that I didn't expect to see is the ensembles / stacking solutions. Sorry if I missed any solutions. I will go through again once more are released and try to consolidate the ones I missed.\n\n**UPDATE**: I have added more model solutions and added tags in the beginning for a quick look up :) If anyone see their solution missing or misinterpreted please leave a comment and I will straighten it out.", "votes": null }, { "id": "1144003", "postDate": "01/08/2021 07:14:19", "content": "

Thank you for creating useful list (and adding mine)!
\nWould you please add some detail about model e.g., SAINT, SAKT, light GBM or ensemble of them?
\nI am interested in which model top rankers used.

", "rawMarkdown": "Thank you for creating useful list (and adding mine)!\nWould you please add some detail about model e.g., SAINT, SAKT, light GBM or ensemble of them?\nI am interested in which model top rankers used.", "votes": null }, { "id": "1144946", "postDate": "01/08/2021 19:05:59", "content": "

@tomooinubushi yup certainly! I will add that today. I will let some more solutions come in before I update this thread but that is a great idea

", "rawMarkdown": "tomooinubushi yup certainly! I will add that today. I will let some more solutions come in before I update this thread but that is a great idea", "votes": null }, { "id": "1145019", "postDate": "01/08/2021 20:11:37", "content": "

16th here: https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798

", "rawMarkdown": "16th here: https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798", "votes": null }, { "id": "1145185", "postDate": "01/09/2021 00:52:57", "content": "

@tomooinubushi please take a look. I have added your suggestion. Let me know what you think :)

", "rawMarkdown": "tomooinubushi please take a look. I have added your suggestion. Let me know what you think :)", "votes": null }, { "id": "1145190", "postDate": "01/09/2021 00:58:01", "content": "

@mpware - great solution! Added 😃

", "rawMarkdown": "mpware - great solution! Added 😃", "votes": null }, { "id": "1145197", "postDate": "01/09/2021 01:08:58", "content": "

Thank you @rdizzl3! It is very informative.
\nThe result is very clear. I realized that I should learn NN models.

", "rawMarkdown": "Thank you @rdizzl3! It is very informative.\nThe result is very clear. I realized that I should learn NN models.", "votes": null }, { "id": "1145686", "postDate": "01/09/2021 09:41:47", "content": "

You can also add our 73th place solution: https://www.kaggle.com/pozitron1997/73th-solution-train-part-single-lgb

", "rawMarkdown": "You can also add our 73th place solution: https://www.kaggle.com/pozitron1997/73th-solution-train-part-single-lgb", "votes": null }, { "id": "1146470", "postDate": "01/09/2021 19:51:38", "content": "

2nd here. https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/210113
\nWe still need to wait for the king.

", "rawMarkdown": "2nd here. https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/210113\nWe still need to wait for the king.", "votes": null } ], "comments": [ { "id": 1144003, "author_name": "tomooinubushi", "author_url": "", "post_date": "01/08/2021 07:14:19", "content": "

Thank you for creating useful list (and adding mine)!
\nWould you please add some detail about model e.g., SAINT, SAKT, light GBM or ensemble of them?
\nI am interested in which model top rankers used.

", "votes": null, "replies": [ { "id": 1144946, "author_name": "rdizzl3", "author_url": "", "post_date": "01/08/2021 19:05:59", "content": "

@tomooinubushi yup certainly! I will add that today. I will let some more solutions come in before I update this thread but that is a great idea

", "votes": null, "replies": [] }, { "id": 1145185, "author_name": "rdizzl3", "author_url": "", "post_date": "01/09/2021 00:52:57", "content": "

@tomooinubushi please take a look. I have added your suggestion. Let me know what you think :)

", "votes": null, "replies": [] }, { "id": 1145197, "author_name": "tomooinubushi", "author_url": "", "post_date": "01/09/2021 01:08:58", "content": "

Thank you @rdizzl3! It is very informative.
\nThe result is very clear. I realized that I should learn NN models.

", "votes": null, "replies": [] } ] }, { "id": 1145019, "author_name": "mpware", "author_url": "", "post_date": "01/08/2021 20:11:37", "content": "

16th here: https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798

", "votes": null, "replies": [ { "id": 1145190, "author_name": "rdizzl3", "author_url": "", "post_date": "01/09/2021 00:58:01", "content": "

@mpware - great solution! Added 😃

", "votes": null, "replies": [] } ] }, { "id": 1145686, "author_name": "pozitron1997", "author_url": "", "post_date": "01/09/2021 09:41:47", "content": "

You can also add our 73th place solution: https://www.kaggle.com/pozitron1997/73th-solution-train-part-single-lgb

", "votes": null, "replies": [] }, { "id": 1146470, "author_name": "mamasinkgs", "author_url": "", "post_date": "01/09/2021 19:51:38", "content": "

2nd here. https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/210113
\nWe still need to wait for the king.

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1143871": "Hey all,\n\nI have made a compilation of the solutions shared so far. I will continue to update as more are released!\n\n2nd [Transformer] - https://www.kaggle.com/mamasinkgs/public-private-2nd-place-solution\n3rd [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209585\n6th [Custom GRU NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209581\n11th [LSTM NN] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209694 \n12th [Stacked Model, LGBM, SAINT, AKT, SSAKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209635\n13th [Stacked Transformer, LSTM, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209622\n16th [Transformer, SAKT, LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798\n17th [Stacked Transfomer, LSTM, GRU] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209713\n18th [Ensemble Catboost, Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209597\n20th [SERT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209587\n22nd [Encoder Modules] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209711\n24th [SAINT++] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209625\n37th [Ensemble SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209624\n38th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209689\n42nd [Ensemble LGB, RAINT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209596\n52nd [Ensemble LGBM, Catboost]- https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209576\n58th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209793\n60th [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209586\n62nd [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209629\n68th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209615\n70th [Transformer] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209584\n71st [LGBM] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209577\n76th [SAINT+] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209599\n125th [Ensemble LGBM, AKT] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209628\n158th [LGBM, FTRL] - https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209619\n\nOne of the things that I find interesting that I didn't expect to see is the ensembles / stacking solutions. Sorry if I missed any solutions. I will go through again once more are released and try to consolidate the ones I missed.\n\n**UPDATE**: I have added more model solutions and added tags in the beginning for a quick look up :) If anyone see their solution missing or misinterpreted please leave a comment and I will straighten it out.", "1144003": "Thank you for creating useful list (and adding mine)!\nWould you please add some detail about model e.g., SAINT, SAKT, light GBM or ensemble of them?\nI am interested in which model top rankers used.", "1144946": "tomooinubushi yup certainly! I will add that today. I will let some more solutions come in before I update this thread but that is a great idea", "1145019": "16th here: https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/209798", "1145185": "tomooinubushi please take a look. I have added your suggestion. Let me know what you think :)", "1145190": "mpware - great solution! Added 😃", "1145197": "Thank you @rdizzl3! It is very informative.\nThe result is very clear. I realized that I should learn NN models.", "1145686": "You can also add our 73th place solution: https://www.kaggle.com/pozitron1997/73th-solution-train-part-single-lgb", "1146470": "2nd here. https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/210113\nWe still need to wait for the king." }, "source": "meta" }