vqc

What problem does it solve?

Classifies tabular data by leveraging a Variational Quantum Classifier (VQC) trained with data re-uploading and the Parameter Shift Rule to produce interpretable class logits from a small quantum circuit.

Core Features & Use Cases

• Data re-uploading encoding of 4 features on 4 qubits per layer to preserve input information across depth.
• Trainable Ry rotations with a CNOT ladder enabling expressive quantum circuits for classification.
• Exact gradient estimation via the Parameter Shift Rule, enabling gradient-based optimization with Adam.
• Demonstrations on the Iris dataset (4 features, 3 classes) and multi-layer quantum-classical training workflows.

Quick Start

Run the included script to train the VQC on Iris data and report the final test accuracy.