{ "id": 368747, "title": "💡How to ensemble predictions -- a key component to every strong solution 🏅", "url": "/competitions/otto-recommender-system/discussion/368747", "author_name": "", "post_date": "2022-11-27T13:55:04.915364100Z", "votes": 12, "comment_count": 2, "views": 0, "content": "

Hey,

\n

Most likely all solutions in the top 50 will involve an ensemble. But how do you ensemble the predictions with over 5 million rows in the submission file, each containing 20 predictions?

\n

I prepared the following notebook to walk you through how to do it 🙂

\n

💡 [2 methods] How-to ensemble predictions 🏅🏅🏅

\n

In it, I show you how to ensemble the predictions even within the constraints of the Kaggle VM. I also introduce two methods of ensembling:

\n\n

Hope this can be of help 🙂 Happy Kaggling!

\n

Other resources you might find useful:

\n", "messages": [ { "id": "2045605", "postDate": "11/27/2022 13:55:04", "content": "

Hey,

\n

Most likely all solutions in the top 50 will involve an ensemble. But how do you ensemble the predictions with over 5 million rows in the submission file, each containing 20 predictions?

\n

I prepared the following notebook to walk you through how to do it 🙂

\n

💡 [2 methods] How-to ensemble predictions 🏅🏅🏅

\n

In it, I show you how to ensemble the predictions even within the constraints of the Kaggle VM. I also introduce two methods of ensembling:

\n\n

Hope this can be of help 🙂 Happy Kaggling!

\n

Other resources you might find useful:

\n", "rawMarkdown": "Hey,\n\nMost likely all solutions in the top 50 will involve an ensemble. But how do you ensemble the predictions with over 5 million rows in the submission file, each containing 20 predictions?\n\nI prepared the following notebook to walk you through how to do it 🙂\n\n[💡 [2 methods] How-to ensemble predictions 🏅🏅🏅](https://www.kaggle.com/code/radek1/2-methods-how-to-ensemble-predictions)\n\nIn it, I show you how to ensemble the predictions even within the constraints of the Kaggle VM. I also introduce two methods of ensembling:\n* vote ensemble\n* vote ensemble with weights\n\nHope this can be of help 🙂 Happy Kaggling!\n\n### Other resources you might find useful:\n\n* [💡 [2 methods] How-to ensemble predictions 🏅🏅🏅](https://www.kaggle.com/code/radek1/2-methods-how-to-ensemble-predictions)\n* [local validation tracks public LB perfecty -- here is the setup](https://www.kaggle.com/competitions/otto-recommender-system/discussion/364991)\n* [💡 For my friends from Twitter and LinkedIn -- here is how to dive into this competition 🐳](https://www.kaggle.com/competitions/otto-recommender-system/discussion/368560)\n* [Full dataset processed to CSV/parquet files with optimized memory footprint](https://www.kaggle.com/competitions/otto-recommender-system/discussion/363843)\n* [co-visitation matrix - simplified, imprvd logic 🔥](https://www.kaggle.com/code/radek1/co-visitation-matrix-simplified-imprvd-logic)\n* [💡 Word2Vec How-to [training and submission]🚀🚀🚀](https://www.kaggle.com/code/radek1/word2vec-how-to-training-and-submission)", "votes": null }, { "id": "2045830", "postDate": "11/27/2022 18:16:13", "content": "

I suggest interested folks can peruse artefacts of the ongoing TPS November 22 challenge. This competition is about ensemble techniques.
\nHope this helps @radek1

", "rawMarkdown": "I suggest interested folks can peruse artefacts of the ongoing TPS November 22 challenge. This competition is about ensemble techniques. \nHope this helps @radek1", "votes": null }, { "id": "2047817", "postDate": "11/29/2022 04:08:44", "content": "

Share a LB boost ensemble notebook.

", "rawMarkdown": "Share a LB boost ensemble notebook.", "votes": null } ], "comments": [ { "id": 2045830, "author_name": "ravi20076", "author_url": "", "post_date": "11/27/2022 18:16:13", "content": "

I suggest interested folks can peruse artefacts of the ongoing TPS November 22 challenge. This competition is about ensemble techniques.
\nHope this helps @radek1

", "votes": null, "replies": [] }, { "id": 2047817, "author_name": "dragonzhang", "author_url": "", "post_date": "11/29/2022 04:08:44", "content": "

Share a LB boost ensemble notebook.

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2045605": "Hey,\n\nMost likely all solutions in the top 50 will involve an ensemble. But how do you ensemble the predictions with over 5 million rows in the submission file, each containing 20 predictions?\n\nI prepared the following notebook to walk you through how to do it 🙂\n\n[💡 [2 methods] How-to ensemble predictions 🏅🏅🏅](https://www.kaggle.com/code/radek1/2-methods-how-to-ensemble-predictions)\n\nIn it, I show you how to ensemble the predictions even within the constraints of the Kaggle VM. I also introduce two methods of ensembling:\n* vote ensemble\n* vote ensemble with weights\n\nHope this can be of help 🙂 Happy Kaggling!\n\n### Other resources you might find useful:\n\n* [💡 [2 methods] How-to ensemble predictions 🏅🏅🏅](https://www.kaggle.com/code/radek1/2-methods-how-to-ensemble-predictions)\n* [local validation tracks public LB perfecty -- here is the setup](https://www.kaggle.com/competitions/otto-recommender-system/discussion/364991)\n* [💡 For my friends from Twitter and LinkedIn -- here is how to dive into this competition 🐳](https://www.kaggle.com/competitions/otto-recommender-system/discussion/368560)\n* [Full dataset processed to CSV/parquet files with optimized memory footprint](https://www.kaggle.com/competitions/otto-recommender-system/discussion/363843)\n* [co-visitation matrix - simplified, imprvd logic 🔥](https://www.kaggle.com/code/radek1/co-visitation-matrix-simplified-imprvd-logic)\n* [💡 Word2Vec How-to [training and submission]🚀🚀🚀](https://www.kaggle.com/code/radek1/word2vec-how-to-training-and-submission)", "2045830": "I suggest interested folks can peruse artefacts of the ongoing TPS November 22 challenge. This competition is about ensemble techniques. \nHope this helps @radek1", "2047817": "Share a LB boost ensemble notebook." }, "source": "meta" }