filmov
tv
[MXML-4-03] Logistic Regression [3/5] - One-vs-Rest (OvR) multiclass classification
Показать описание
In the last two videos, we looked at the binary classification. In this video, we will look at the multiclass classification.
First, let's look at the OvR, One-vs-Rest method. And let's implement this using scikit-learn's LogisticRegression library.
One-vs-Rest is a heuristic method that uses multiple binary classification models for multiclass classification, shortly called OvR. OvR is also called OvA or One versus All.
Another method is Multinomial or Softmax Logistic Regression. This is an algorithmic method that uses the softmax function and the cross entropy loss function. We'll take a closer look at this method in the next video.
#LogisticRegression #MulticlassClassification #One-vs-Rest #OvR #One-vs-All #OvA
First, let's look at the OvR, One-vs-Rest method. And let's implement this using scikit-learn's LogisticRegression library.
One-vs-Rest is a heuristic method that uses multiple binary classification models for multiclass classification, shortly called OvR. OvR is also called OvA or One versus All.
Another method is Multinomial or Softmax Logistic Regression. This is an algorithmic method that uses the softmax function and the cross entropy loss function. We'll take a closer look at this method in the next video.
#LogisticRegression #MulticlassClassification #One-vs-Rest #OvR #One-vs-All #OvA