Overfitting in quantum machine learning and entangling dropout
This work addresses overfitting for quantum machine learning practitioners, but it is incremental as it adapts a classical technique to the quantum domain.
The paper tackles overfitting in quantum machine learning by proposing entangling dropout, a method that randomly removes entangling gates during training to reduce circuit expressibility, and demonstrates through simple case studies that it effectively suppresses overfitting.
The ultimate goal in machine learning is to construct a model function that has a generalization capability for unseen dataset, based on given training dataset. If the model function has too much expressibility power, then it may overfit to the training data and as a result lose the generalization capability. To avoid such overfitting issue, several techniques have been developed in the classical machine learning regime, and the dropout is one such effective method. This paper proposes a straightforward analogue of this technique in the quantum machine learning regime, the entangling dropout, meaning that some entangling gates in a given parametrized quantum circuit are randomly removed during the training process to reduce the expressibility of the circuit. Some simple case studies are given to show that this technique actually suppresses the overfitting.