I have a hybrid data that contains 15 categorical data and 4 continuous data. I need to implement a prediction on the data. So as I don't have any labeled data, I need to implement the unsupervised algorithm.
Now Using PCA my dimensions reduced to 5 but still it has 3 categorical and 2 continuous data. Now to create a prediction algorithm, I can use KMeans for continuous data or KModes for categorical data, But I couldn't find any algorithm that works on both types of data.
Can anyone know any algorithm that works on both the data types for unsupervised prediction?