How to classify objects with different number of features?

Question

Machine learning

Good afternoon. Somehow, you need to classify objects with a different number of features (based on the structure of the object itself). I decided to practice with klearn in Python at the same time. Can anyone suggest which classification algorithm to use without knowing the number of features?

Answer 1

You can use any one, bringing the data to the same dimension - for example, fill the gaps in the input matrix with zeros, a separate None value, the most common value, etc.
You can use variable-dimensional input data, for example, in LSTM networks, but it is better not to take on such tricks without some experience.