Data that is used for classification but having to class label or target variables. Then how we set its target variable ?
https://stackoverflow.com/questions/44821978/how-to-build-target-variable-for-supervised-machine-learning-project
https://stackoverflow.com/questions/44821978/how-to-build-target-variable-for-supervised-machine-learning-project
Principal Component Analysis is also capable of giving you some hints about unknown classes. Plotting scores of observations on first couple of PC's (which explain most of the variance) is beneficial along with EDA in Paramjit Kour 's link.
Use PCA with all variables. Extract PC1 (and maybe PC2 if contributes to explained variance enough). When you check scores of observations on principal components, you can possibly see some of them are close to each other (with more than one group) and some are outliers. Then, you can label them as classes and use this information for test data.
- Badges
- Science topic