I have a dataset that consists of a target column with 65 classes. Also, the dataset has 200 columns/features. I researched multi-label classification and found the popular algorithms that can be used for multi-class classification include:
- k-Nearest Neighbors
- Decision Trees
- Naive Bayes
- Random Forest
- Gradient Boosting etc.,
However, I saw multi-label classification examples that have four or five target categories.
Also from the Scikit-Learn Documentation on Multi-class:
|Number of targets||Target cardinality||Valid type_of_target|
|Multilabel classification||>1||2 (0 or 1)||‘multilabel-indicator’|
Found no maximum limit for cardinality in Multi-class. Is there any cardinality limit for Multi-class classification? Are there any popular algorithms that can be used for predicting the classes that contain more than 65 classes?