I currently use GridSearch.CV to find the best possible parameters for the XGBoost Algorithm. Is there any other way?
", "messages": [ { "id": 1916115, "postDate": "2022-08-27T16:37:38.463Z", "content": "I currently use GridSearch.CV to find the best possible parameters for the XGBoost Algorithm. Is there any other way?
", "rawMarkdown": "I currently use GridSearch.CV to find the best possible parameters for the XGBoost Algorithm. Is there any other way?", "votes": 1 }, { "id": 1917270, "postDate": "2022-08-28T15:36:23.703Z", "content": "Gridsearch CV is very computationaly expensive especially for this competition in which size of dataset is large. You can either use Optuna or Random search CV. My personal favorite is optuna as it provide many good features over Random or Grid search CV
", "rawMarkdown": "Gridsearch CV is very computationaly expensive especially for this competition in which size of dataset is large. You can either use Optuna or Random search CV. My personal favorite is optuna as it provide many good features over Random or Grid search CV", "votes": 2, "replies": [ { "id": 1919465, "postDate": "2022-08-30T12:42:23.383Z", "content": "I agree with @guptadikshant; I think Optuna is your best option on this dataset.
", "rawMarkdown": "I agree with @guptadikshant; I think Optuna is your best option on this dataset." } ] }, { "id": 1916314, "postDate": "2022-08-27T18:50:50.563Z", "content": "For reference, this topic has an over-abundance of information to sift through: https://en.wikipedia.org/wiki/Hyperparameter_optimization
\nIn practice, I'm also curious what people actually use. I've certainly heard Optuna mentioned before, and this notebook is a fully working(?) example of BayesianOptimization: https://www.kaggle.com/code/bogorodvo/bayesianoptimization-p-2-lst-stratification
", "rawMarkdown": "For reference, this topic has an over-abundance of information to sift through: https://en.wikipedia.org/wiki/Hyperparameter_optimization\n\nIn practice, I'm also curious what people actually use. I've certainly heard Optuna mentioned before, and this notebook is a fully working(?) example of BayesianOptimization: https://www.kaggle.com/code/bogorodvo/bayesianoptimization-p-2-lst-stratification " }, { "id": 1916189, "postDate": "2022-08-27T17:20:52.263Z", "content": "I presume one could try and use optuna for the same
", "rawMarkdown": "I presume one could try and use optuna for the same" }, { "id": 1916119, "postDate": "2022-08-27T16:41:16.567Z", "content": "optuna or randomgridsearch works better than GridSearch @harshm27
", "rawMarkdown": "optuna or randomgridsearch works better than GridSearch @harshm27 " }, { "id": 1921356, "postDate": "2022-08-31T19:01:10.703Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 1917270, "author_name": "Dikshant Gupta", "author_url": "", "post_date": "2022-08-28T15:36:23.703000", "content": "Gridsearch CV is very computationaly expensive especially for this competition in which size of dataset is large. You can either use Optuna or Random search CV. My personal favorite is optuna as it provide many good features over Random or Grid search CV
", "votes": 2, "replies": [ { "id": 1919465, "author_name": "Oscar Aguilar", "author_url": "", "post_date": "2022-08-30T12:42:23.383000", "content": "I agree with @guptadikshant; I think Optuna is your best option on this dataset.
", "votes": 0, "replies": [] } ] }, { "id": 1916314, "author_name": "Robert Hatch", "author_url": "", "post_date": "2022-08-27T18:50:50.563000", "content": "For reference, this topic has an over-abundance of information to sift through: https://en.wikipedia.org/wiki/Hyperparameter_optimization
\nIn practice, I'm also curious what people actually use. I've certainly heard Optuna mentioned before, and this notebook is a fully working(?) example of BayesianOptimization: https://www.kaggle.com/code/bogorodvo/bayesianoptimization-p-2-lst-stratification
", "votes": 0, "replies": [] }, { "id": 1916189, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2022-08-27T17:20:52.263000", "content": "I presume one could try and use optuna for the same
", "votes": 0, "replies": [] }, { "id": 1916119, "author_name": "Ram Jas", "author_url": "", "post_date": "2022-08-27T16:41:16.567000", "content": "optuna or randomgridsearch works better than GridSearch @harshm27
", "votes": 0, "replies": [] }, { "id": 1921356, "author_name": "", "author_url": "", "post_date": "2022-08-31T19:01:10.703000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1916115": "I currently use GridSearch.CV to find the best possible parameters for the XGBoost Algorithm. Is there any other way?", "1917270": "Gridsearch CV is very computationaly expensive especially for this competition in which size of dataset is large. You can either use Optuna or Random search CV. My personal favorite is optuna as it provide many good features over Random or Grid search CV", "1916314": "For reference, this topic has an over-abundance of information to sift through: https://en.wikipedia.org/wiki/Hyperparameter_optimization\n\nIn practice, I'm also curious what people actually use. I've certainly heard Optuna mentioned before, and this notebook is a fully working(?) example of BayesianOptimization: https://www.kaggle.com/code/bogorodvo/bayesianoptimization-p-2-lst-stratification ", "1916189": "I presume one could try and use optuna for the same", "1916119": "optuna or randomgridsearch works better than GridSearch @harshm27 ", "1921356": "" } }