I am trying to perform a classification task with qiskit's VQC. The dataset I am using has a large number of dimensions/features/columns. I am trying to figure out which feature map works best. Also, I only found four feature maps in qiskit documentation, three under Data Encoding Circuits and one here, are there more available ?
Without dimentionality reduction is there any other way of encoding more features on fewer qubits (like the RawFeatureVector) ?
How to theoretically select the best feature map ?