{ "id": 322863, "title": "I like this new word: emsembling", "url": "/competitions/h-and-m-personalized-fashion-recommendations/discussion/322863", "author_name": "Gabriel Preda", "post_date": "2022-05-04T06:33:53.984000", "votes": 17, "comment_count": 7, "views": 0, "content": "

I like this new word, ensembling.
\nIt sounds like combining ensemble + gambling.

\n

", "messages": [ { "id": 1776704, "postDate": "2022-05-04T06:33:53.983Z", "content": "

I like this new word, ensembling.
\nIt sounds like combining ensemble + gambling.

\n

", "rawMarkdown": "I like this new word, ensembling.\nIt sounds like combining ensemble + gambling.\n\n", "votes": 18 }, { "id": 1778810, "postDate": "2022-05-05T17:41:28.307Z", "content": "

It seems that ensembling does not perform well in time-series, quantitative competitions. The top 1% in other competitions build on the ensembling while relying on a change in algorithmic direction, innovation, or deep business understanding.

", "rawMarkdown": "It seems that ensembling does not perform well in time-series, quantitative competitions. The top 1% in other competitions build on the ensembling while relying on a change in algorithmic direction, innovation, or deep business understanding.", "votes": 3, "replies": [ { "id": 1778815, "postDate": "2022-05-05T17:47:00.630Z", "content": "

Agreed. Business sense is important.

", "rawMarkdown": "Agreed. Business sense is important.", "votes": 3 } ] }, { "id": 1779289, "postDate": "2022-05-06T08:53:06.553Z", "content": "

Most high-scoring ensembles are gambling, aren't they?

", "rawMarkdown": "Most high-scoring ensembles are gambling, aren't they?\n", "votes": 1 }, { "id": 1776716, "postDate": "2022-05-04T06:41:13.507Z", "content": "

Dear @gpreda

\n

I am sure as you well know, ensembling does not at all have to be synonymous with gambling; I think one of the great milestones in ML was the transition from the decision tree (a 'weak learner') to the random forest (an ensemble of decision trees), collectively leading to a strong learner.

\n

All the best,
\ncarl

", "rawMarkdown": "Dear @gpreda \n\nI am sure as you well know, ensembling does not at all have to be synonymous with gambling; I think one of the great milestones in ML was the transition from the decision tree (a 'weak learner') to the random forest (an ensemble of decision trees), collectively leading to a strong learner. \n\nAll the best,\ncarl ", "votes": 2, "replies": [ { "id": 1776720, "postDate": "2022-05-04T06:44:24.520Z", "content": "

I know but I like the sound of it

", "rawMarkdown": "I know but I like the sound of it", "votes": 6 } ] }, { "id": 1780124, "postDate": "2022-05-07T05:35:26.490Z", "content": "

Seems like multi-task gambling lol

", "rawMarkdown": "Seems like multi-task gambling lol" }, { "id": 1778827, "postDate": "2022-05-05T17:55:10.520Z", "rawMarkdown": "", "votes": 1, "isDeleted": true } ], "comments": [ { "id": 1778810, "author_name": "Kz", "author_url": "", "post_date": "2022-05-05T17:41:28.307000", "content": "

It seems that ensembling does not perform well in time-series, quantitative competitions. The top 1% in other competitions build on the ensembling while relying on a change in algorithmic direction, innovation, or deep business understanding.

", "votes": 3, "replies": [ { "id": 1778815, "author_name": "Heartquake", "author_url": "", "post_date": "2022-05-05T17:47:00.630000", "content": "

Agreed. Business sense is important.

", "votes": 3, "replies": [] } ] }, { "id": 1779289, "author_name": "VVVVX", "author_url": "", "post_date": "2022-05-06T08:53:06.553000", "content": "

Most high-scoring ensembles are gambling, aren't they?

", "votes": 1, "replies": [] }, { "id": 1776716, "author_name": "Carl McBride Ellis", "author_url": "", "post_date": "2022-05-04T06:41:13.507000", "content": "

Dear @gpreda

\n

I am sure as you well know, ensembling does not at all have to be synonymous with gambling; I think one of the great milestones in ML was the transition from the decision tree (a 'weak learner') to the random forest (an ensemble of decision trees), collectively leading to a strong learner.

\n

All the best,
\ncarl

", "votes": 2, "replies": [ { "id": 1776720, "author_name": "Gabriel Preda", "author_url": "", "post_date": "2022-05-04T06:44:24.520000", "content": "

I know but I like the sound of it

", "votes": 6, "replies": [] } ] }, { "id": 1780124, "author_name": "Zelin", "author_url": "", "post_date": "2022-05-07T05:35:26.490000", "content": "

Seems like multi-task gambling lol

", "votes": 0, "replies": [] }, { "id": 1778827, "author_name": "", "author_url": "", "post_date": "2022-05-05T17:55:10.520000", "content": "", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "1776704": "I like this new word, ensembling.\nIt sounds like combining ensemble + gambling.\n\n", "1778810": "It seems that ensembling does not perform well in time-series, quantitative competitions. The top 1% in other competitions build on the ensembling while relying on a change in algorithmic direction, innovation, or deep business understanding.", "1779289": "Most high-scoring ensembles are gambling, aren't they?\n", "1776716": "Dear @gpreda \n\nI am sure as you well know, ensembling does not at all have to be synonymous with gambling; I think one of the great milestones in ML was the transition from the decision tree (a 'weak learner') to the random forest (an ensemble of decision trees), collectively leading to a strong learner. \n\nAll the best,\ncarl ", "1780124": "Seems like multi-task gambling lol", "1778827": "" } }