Abstract
Quantum machine learning is a promising application of quantum computing for data classification. However, most of the previous research focused on binary classification, and there are few studies on multi-classification. The major challenge comes from the limitations of near-term quantum devices on the number of qubits and the size of quantum circuits. In this paper, we propose a hybrid quantum neural network to implement multi-classification of a real-world dataset. We use an average pooling downsampling strategy to reduce the dimensionality of samples, and we design a ladder-like parameterized quantum circuit to disentangle the input states. Besides this, we adopt an all-qubit multi-observable measurement strategy to capture sufficient hidden information from the quantum system. The experimental results show that our algorithm outperforms the classical neural network and performs especially well on different multi-class datasets, which provides some enlightenment for the application of quantum computing to real-world data on near-term quantum processors.
Highlights
Quantum computers can take advantage of the superposition and entanglement of quantum systems to be exponentially faster than classical computers on certain computing tasks [1]
Machine learning is an attractive application of quantum computers, because it has made great progress in solving complex practical tasks on classical computers, and because its inherent noise resistance is beneficial for realization on Noisy Intermediate-Scale Quantum (NISQ) devices without error correction
It is theoretically possible that the classical data can be prepared in a quantum superposition by quantum random access memory (QRAM), its physical realization will require larger quantum computers in the future
Summary
Quantum computers can take advantage of the superposition and entanglement of quantum systems to be exponentially faster than classical computers on certain computing tasks [1]. Li et al [27] investigated a quantum deep convolutional neural network (QDCNN) model and employed MNIST and GTSRB datasets for 10-way classification As they mentioned, it is theoretically possible that the classical data can be prepared in a quantum superposition by quantum random access memory (QRAM), its physical realization will require larger quantum computers in the future. In the process of dimensionality reduction, the average pooling downsampling strategy is adopted to compress the image; in the quantum part, first we use a quantum encoding circuit to prepare the initial quantum state, we apply a parameterized quantum circuit (PQC) to perform unitary transformations on this quantum state, and, we perform quantum measurements to calculate the expectation values of observables; in the classical part, we employ the classical neural network to post-process the outcomes of the quantum part, and we use a classical optimizer to adjust both the quantum and classical parameters. Thpeamrtiidsdthleepmaertasisuroenmeelnayt eorpoerfaltaidond,ewr-lhikicehPcQalCcuwlahteicshthreeqeuxpireecsta(ntio−n1v) aplaureasmoef taersse.t Tohf eobrisgehrvt-ahbalensdon paertacishtqhuebmit.easurement operation, which calculates the expectation values of a set of observables on each qubit
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
