[MXML-8-03] Random Forest [3/7] - Out-Of-Bag (OOB) score

* This video was produced in Korean and translated into English. And the voice is generated by AI TTS. The English translation may contain grammatical errors.
This is part 3 of a series on Random Forest.
In this video, we will talk about the OOB score, which is the Out-Of-Bag score. We'll take a look at the OOB through an example and implement it in a code. We'll do this from scratch as we did in the previous video, and we'll also do it with scikit-learn.
Data points that are not selected by row subsampling are called Out-Of-Bag, OOB data. Conversely, the selected data points are called In-Of-Bag, IOB data. OOB data can be used to evaluate the performance of a model.
IOB data is used to create a model, while OOB data is used to evaluate the model. The error rate measured by OOB is called the OOB score. When using OOB score, no cross-validation is required. This means there is no need to create separate validation data. The OOB data can be used instead of validation data. The OOB score can determine the optimal tree depth, optimal number of tree models, etc.
If you have large enough data, you can create and use a separate validation dataset. However, if you have less data, you can use OOB data for model validation without validation data. OOB data allows you to use your data more efficiently.
#RandomForest #OutOfBag #OOB #InOfBag #IOB #OOBScore

[@meanxai]

The code can be found in github.com/meanxai/machine_learning .