Well, my best one is in the competition Titanic with predicting everyone dead.😂
", "rawMarkdown": "Well, my best one is in the competition Titanic with predicting everyone dead.😂 ", "votes": 4 }, { "id": 698745, "postDate": "2019-12-19T16:51:08.877Z", "content": "I see many of us have lower than 1.06 without leak which I think is very good results.
", "rawMarkdown": "I see many of us have lower than 1.06 without leak which I think is very good results. ", "votes": 2 }, { "id": 697635, "postDate": "2019-12-18T07:52:23.660Z", "content": "1.024 blend of 3 models.
", "rawMarkdown": "1.024 blend of 3 models.", "votes": 2, "replies": [ { "id": 697652, "postDate": "2019-12-18T08:25:23.620Z", "content": "cool!
", "rawMarkdown": "cool! ", "votes": 1 }, { "id": 697671, "postDate": "2019-12-18T08:54:30.440Z", "content": "Thanks, it's mostly my teammate's work.
", "rawMarkdown": "Thanks, it's mostly my teammate's work." }, { "id": 697967, "postDate": "2019-12-18T15:53:55.377Z", "content": "that's really impressive
", "rawMarkdown": "that's really impressive", "votes": 1 }, { "id": 697973, "postDate": "2019-12-18T15:59:39.997Z", "content": "Thanks. After shake-up it probably will not be!
", "rawMarkdown": "Thanks. After shake-up it probably will not be!" }, { "id": 698010, "postDate": "2019-12-18T17:01:14.343Z", "content": "You think so? I think it will, since shakeup will be mostly due to leaks not being in private LB
", "rawMarkdown": "You think so? I think it will, since shakeup will be mostly due to leaks not being in private LB" }, { "id": 698472, "postDate": "2019-12-19T09:36:00.910Z", "content": "Sure. But it's more about the structure of data which reminds me of the LANL competition.
", "rawMarkdown": "Sure. But it's more about the structure of data which reminds me of the LANL competition." } ] }, { "id": 698532, "postDate": "2019-12-19T11:01:20.287Z", "content": "1.060 single model. 1.040 blend of 3 models.
", "rawMarkdown": "1.060 single model. 1.040 blend of 3 models.", "votes": 1 }, { "id": 690621, "postDate": "2019-12-08T21:29:39.673Z", "content": "LB 1.07, Single model.
", "rawMarkdown": "LB 1.07, Single model.", "votes": 1, "replies": [ { "id": 691086, "postDate": "2019-12-09T15:20:28.393Z", "content": "upd: 1.06 single model
", "rawMarkdown": "upd: 1.06 single model" } ] }, { "id": 701949, "postDate": "2019-12-24T05:08:10.527Z", "content": "Preliminary 1.138 with single model. But I still can't see the results. Does it mean they block me out?
", "rawMarkdown": "Preliminary 1.138 with single model. But I still can't see the results. Does it mean they block me out?" }, { "id": 697809, "postDate": "2019-12-18T12:50:20.947Z", "content": "1.066 single model
", "rawMarkdown": "1.066 single model", "replies": [ { "id": 698741, "postDate": "2019-12-19T16:46:28.660Z", "content": "upd: 1.061 single model
", "rawMarkdown": "upd: 1.061 single model" } ] }, { "id": 697147, "postDate": "2019-12-17T14:33:38.470Z", "content": "1.063 single, 1.057 blend with 3 public kernels and 0.55 weight for my model
", "rawMarkdown": "1.063 single, 1.057 blend with 3 public kernels and 0.55 weight for my model" }, { "id": 690750, "postDate": "2019-12-09T04:14:55.363Z", "content": "1.06 so far. I lost interest a far ago... I have not done almost anything currently XD. However, It is enough to learn something from the competition. That was my initial goal.
", "rawMarkdown": "1.06 so far. I lost interest a far ago... I have not done almost anything currently XD. However, It is enough to learn something from the competition. That was my initial goal." }, { "id": 689968, "postDate": "2019-12-07T18:50:02.153Z", "content": "1.06 - single model
", "rawMarkdown": "1.06 - single model " }, { "id": 689805, "postDate": "2019-12-07T12:57:58.323Z", "content": "1.06 single model. UP: 1.04 blending
", "rawMarkdown": "1.06 single model. UP: 1.04 blending" }, { "id": 687977, "postDate": "2019-12-05T04:47:57.510Z", "content": "1.06 LGB
", "rawMarkdown": "1.06 LGB" }, { "id": 685567, "postDate": "2019-12-01T23:38:17.860Z", "content": "A similar post with the same question here.
", "rawMarkdown": "A similar post with the same question [here](https://www.kaggle.com/c/ashrae-energy-prediction/discussion/118214)." }, { "id": 685090, "postDate": "2019-12-01T00:30:02.510Z", "content": "1.05 -- lots of internal blending of my own models, but no external data (outside of time-zone information).
", "rawMarkdown": "1.05 -- lots of internal blending of my own models, but no external data (outside of time-zone information).", "replies": [ { "id": 685145, "postDate": "2019-12-01T03:17:45.767Z", "content": "nice score, our lb is 1.06 without leak data
", "rawMarkdown": "nice score, our lb is 1.06 without leak data", "votes": 1 }, { "id": 685152, "postDate": "2019-12-01T03:32:11.233Z", "content": "Well, you must be very effective when applying leak data, then, because y'all are still managing to hold an 11 position lead on me. It'll be fascinating to see how it all shakes out on the private LB.
", "rawMarkdown": "Well, you must be very effective when applying leak data, then, because y'all are still managing to hold an 11 position lead on me. It'll be fascinating to see how it all shakes out on the private LB." } ] }, { "id": 701254, "postDate": "2019-12-23T09:41:49.533Z", "rawMarkdown": "", "votes": -1, "isDeleted": true } ], "comments": [ { "id": 684795, "author_name": "Zijian Zhang", "author_url": "", "post_date": "2019-11-30T13:41:31.437000", "content": "Well, my best one is in the competition Titanic with predicting everyone dead.😂
", "votes": 4, "replies": [] }, { "id": 698745, "author_name": "Mukharbek Organokov", "author_url": "", "post_date": "2019-12-19T16:51:08.877000", "content": "I see many of us have lower than 1.06 without leak which I think is very good results.
", "votes": 2, "replies": [] }, { "id": 697635, "author_name": "Sia", "author_url": "", "post_date": "2019-12-18T07:52:23.660000", "content": "1.024 blend of 3 models.
", "votes": 2, "replies": [ { "id": 697652, "author_name": "BestPredict", "author_url": "", "post_date": "2019-12-18T08:25:23.620000", "content": "cool!
", "votes": 1, "replies": [] }, { "id": 697671, "author_name": "Sia", "author_url": "", "post_date": "2019-12-18T08:54:30.440000", "content": "Thanks, it's mostly my teammate's work.
", "votes": 0, "replies": [] }, { "id": 697967, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-18T15:53:55.377000", "content": "that's really impressive
", "votes": 1, "replies": [] }, { "id": 697973, "author_name": "Sia", "author_url": "", "post_date": "2019-12-18T15:59:39.997000", "content": "Thanks. After shake-up it probably will not be!
", "votes": 0, "replies": [] }, { "id": 698010, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-18T17:01:14.343000", "content": "You think so? I think it will, since shakeup will be mostly due to leaks not being in private LB
", "votes": 0, "replies": [] }, { "id": 698472, "author_name": "Sia", "author_url": "", "post_date": "2019-12-19T09:36:00.910000", "content": "Sure. But it's more about the structure of data which reminds me of the LANL competition.
", "votes": 0, "replies": [] } ] }, { "id": 698532, "author_name": "FISH", "author_url": "", "post_date": "2019-12-19T11:01:20.287000", "content": "1.060 single model. 1.040 blend of 3 models.
", "votes": 1, "replies": [] }, { "id": 690621, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-08T21:29:39.673000", "content": "LB 1.07, Single model.
", "votes": 1, "replies": [ { "id": 691086, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-09T15:20:28.393000", "content": "upd: 1.06 single model
", "votes": 0, "replies": [] } ] }, { "id": 701949, "author_name": "Irina Chuchueva", "author_url": "", "post_date": "2019-12-24T05:08:10.527000", "content": "Preliminary 1.138 with single model. But I still can't see the results. Does it mean they block me out?
", "votes": 0, "replies": [] }, { "id": 697809, "author_name": "Dean", "author_url": "", "post_date": "2019-12-18T12:50:20.947000", "content": "1.066 single model
", "votes": 0, "replies": [ { "id": 698741, "author_name": "Dean", "author_url": "", "post_date": "2019-12-19T16:46:28.660000", "content": "upd: 1.061 single model
", "votes": 0, "replies": [] } ] }, { "id": 697147, "author_name": "Iurii Uspangaliev", "author_url": "", "post_date": "2019-12-17T14:33:38.470000", "content": "1.063 single, 1.057 blend with 3 public kernels and 0.55 weight for my model
", "votes": 0, "replies": [] }, { "id": 690750, "author_name": "Hanjoon Choe", "author_url": "", "post_date": "2019-12-09T04:14:55.363000", "content": "1.06 so far. I lost interest a far ago... I have not done almost anything currently XD. However, It is enough to learn something from the competition. That was my initial goal.
", "votes": 0, "replies": [] }, { "id": 689968, "author_name": "AK", "author_url": "", "post_date": "2019-12-07T18:50:02.153000", "content": "1.06 - single model
", "votes": 0, "replies": [] }, { "id": 689805, "author_name": "MPWARE", "author_url": "", "post_date": "2019-12-07T12:57:58.323000", "content": "1.06 single model. UP: 1.04 blending
", "votes": 0, "replies": [] }, { "id": 687977, "author_name": "Kamal Chhirang", "author_url": "", "post_date": "2019-12-05T04:47:57.510000", "content": "1.06 LGB
", "votes": 0, "replies": [] }, { "id": 685567, "author_name": "Mohammed Tayor", "author_url": "", "post_date": "2019-12-01T23:38:17.860000", "content": "A similar post with the same question here.
", "votes": 0, "replies": [] }, { "id": 685090, "author_name": "Robert Stockton", "author_url": "", "post_date": "2019-12-01T00:30:02.510000", "content": "1.05 -- lots of internal blending of my own models, but no external data (outside of time-zone information).
", "votes": 0, "replies": [ { "id": 685145, "author_name": "BestPredict", "author_url": "", "post_date": "2019-12-01T03:17:45.767000", "content": "nice score, our lb is 1.06 without leak data
", "votes": 1, "replies": [] }, { "id": 685152, "author_name": "Robert Stockton", "author_url": "", "post_date": "2019-12-01T03:32:11.233000", "content": "Well, you must be very effective when applying leak data, then, because y'all are still managing to hold an 11 position lead on me. It'll be fascinating to see how it all shakes out on the private LB.
", "votes": 0, "replies": [] } ] }, { "id": 701254, "author_name": "", "author_url": "", "post_date": "2019-12-23T09:41:49.533000", "content": "", "votes": -1, "replies": [] } ], "raw_markdown_by_id": { "684789": "", "684795": "Well, my best one is in the competition Titanic with predicting everyone dead.😂 ", "698745": "I see many of us have lower than 1.06 without leak which I think is very good results. ", "697635": "1.024 blend of 3 models.", "698532": "1.060 single model. 1.040 blend of 3 models.", "690621": "LB 1.07, Single model.", "701949": "Preliminary 1.138 with single model. But I still can't see the results. Does it mean they block me out?", "697809": "1.066 single model", "697147": "1.063 single, 1.057 blend with 3 public kernels and 0.55 weight for my model", "690750": "1.06 so far. I lost interest a far ago... I have not done almost anything currently XD. However, It is enough to learn something from the competition. That was my initial goal.", "689968": "1.06 - single model ", "689805": "1.06 single model. UP: 1.04 blending", "687977": "1.06 LGB", "685567": "A similar post with the same question [here](https://www.kaggle.com/c/ashrae-energy-prediction/discussion/118214).", "685090": "1.05 -- lots of internal blending of my own models, but no external data (outside of time-zone information).", "701254": "" } }