Interface administrators, Administrators (Semantic MediaWiki), Curators (Semantic MediaWiki), Editors (Semantic MediaWiki), Suppressors, Administrators
7,785
edits
m (Text replacement - "Category:Machine learning terms" to "Category:Machine learning terms Category:not updated") |
|||
(One intermediate revision by the same user not shown) | |||
Line 14: | Line 14: | ||
#[[Cross-validation]]: This method is commonly used for evaluating model stability. In this approach, data is divided into multiple [[fold]]s and the model trained and assessed on each fold. The average performance across all folds then serves to judge how stable the model truly is. | #[[Cross-validation]]: This method is commonly used for evaluating model stability. In this approach, data is divided into multiple [[fold]]s and the model trained and assessed on each fold. The average performance across all folds then serves to judge how stable the model truly is. | ||
#[[Bootstrapping]]: This resampling technique involves drawing multiple examples with replacement from the training data to generate multiple [[training set | #[[Bootstrapping]]: This resampling technique involves drawing multiple examples with replacement from the training data to generate multiple [[training set]]s. The model is then trained on each set and its average performance used to assess its stability. | ||
#[[Regularization]]: This technique helps control [[overfitting]] in a model by adding a penalty term to the loss function. Regularization helps improve model stability by preventing it from fitting [[noise]] in data. | #[[Regularization]]: This technique helps control [[overfitting]] in a model by adding a penalty term to the loss function. Regularization helps improve model stability by preventing it from fitting [[noise]] in data. | ||
Line 20: | Line 20: | ||
Stability is a model's capacity to remain accurate even when small changes are made to its training data or construction. Think of it like building a tower with blocks; if it is constructed poorly, even minor winds could bring it down; on the other hand, if built robustly and securely, even strong winds won't knock it over. We can make our tower stronger using techniques like cross-validation, bootstrapping, and regularization. | Stability is a model's capacity to remain accurate even when small changes are made to its training data or construction. Think of it like building a tower with blocks; if it is constructed poorly, even minor winds could bring it down; on the other hand, if built robustly and securely, even strong winds won't knock it over. We can make our tower stronger using techniques like cross-validation, bootstrapping, and regularization. | ||
[[Category:Terms]] [[Category:Machine learning terms]] | [[Category:Terms]] [[Category:Machine learning terms]] [[Category:not updated]] |