Search

Bayesian networks are powerful tools for probabilistic analysis and have been widely used in machine learning and data science. Unlike the time-consuming parameter training process of neural networks, Bayes classifiers constructed on Bayesian networks can make decisions based solely on statistical data from samples. In this paper, we focus on constructing quantum Bayes classifiers (QBCs). We design both a naive QBC and three semi-naive QBCs (SN-QBCs). These QBCs are then applied to image classification tasks. To reduce computational complexity, we employ a local feature sampling method to extract a limited number of feature attributes from an image. These attributes serve as nodes of the Bayesian networks to generate the QBCs. We simulate these QBCs on the MindQuantum platform and evaluate their performance on the MNIST and Fashion-MNIST datasets. Our results demonstrate that these QBCs achieve good classification accuracies even with a limited number of attributes. The classification accuracies of QBCs on the MNIST dataset surpass those of classical Bayesian networks and quantum neural networks that utilize all available feature attributes.

Quantum machine learning has shown advantages in many ways compared to classical machine learning. In machine learning, a difficult problem is how to learn a model with high robustness and strong generalization ability from a limited feature space. Combining multiple models as base learners, ensemble learning (EL) can effectively improve the accuracy, generalization ability, and robustness of the final model. The key to EL lies in two aspects, the performance of base learners and the choice of the combination strategy. Recently, quantum EL (QEL) has been studied. However, existing combination strategies in QEL are inadequate in considering the accuracy and variance among base learners. This paper presents a hybrid EL framework that combines quantum and classical advantages. More importantly, we propose an efficient combination strategy for improving the accuracy of classification in the framework. We verify the feasibility and efficiency of our framework and strategy by using the MNIST dataset. Simulation results show that the hybrid EL framework with our combination strategy not only has a higher accuracy and lower variance than the single model without the ensemble, but also has a better accuracy than the majority voting and the weighted voting strategies in most cases.

Reconstruction of large-sized data from small-sized ones is an important problem in information science, and a typical example is the image super-resolution reconstruction in computer vision. Combining machine learning and quantum computing, quantum machine learning has shown the ability to accelerate data processing and provides new methods for information processing. In this paper, we propose two frameworks for data reconstruction based on quantum neural networks (QNNs) and quantum autoencoder (QAE). The effects of the two frameworks are evaluated by using the MNIST handwritten digits as datasets. Simulation results show that QNNs and QAE can work well for data reconstruction. We also compare our results with classical super-resolution neural networks, and the results of one QNN are very close to classical ones.

Quantum cryptography has attracted much attention in recent years. In most existing quantum cryptographic protocols, players usually need the full quantum power of generating, manipulating or measuring quantum states. Semiquantum cryptography was proposed to deal with the issue that some players require only partial quantum power, such as preparing or measuring quantum states in the classical basis, which simplifies the implementations of quantum cryptography. However, the efficiency of the existing semiquantum cryptographic protocols was relatively low from a practical point of view. In this paper, we devise some new semiquantum key distribution (SQKD) protocols which highly improve the efficiency of the most well-known SQKD protocols [Phys. Rev. Lett. 99, 140501 (2007) & Phys. Rev. A 79, 052312 (2009)]. By letting players select their actions asymmetrically, the efficiency of our new protocols can be made asymptotically close to 100%. Besides, one of our proposed protocols also utilizes the discarded X-SIFT bits in the original SQKD protocol, which further improves the efficiency of SQKD. We prove that the proposed SQKD protocols are completely robust against the most general attack.

Quantum secure communication provides a new way for protecting the security of information. As an important component of quantum secure communication, remote state preparation (RSP) can securely transmit a quantum state from a sender to a remote receiver. The existence of quantum noise severely affects the security and reliability of quantum communication system. In this paper, we study the method for improving the efficiency of joint RSP (JRSP) subjected to noise with the help of weak measurement and its reversal measurement. Taking a GHZ based deterministic JRSP as an example, we utilize the technique of weak measurement and its reversal to suppress the effect of the amplitude-damping noise firstly. Our study shows that the fidelity of the output state can be improved in the amplitude-damping noise. We also study the effect of weak measurement and its reversal in other three types of noise usually encountered in real-world, namely, the bit-flip, phase-flip (phase-damping) and depolarizing noise. Our results show that the weak measurement has no effect for suppressing the bit-flip and phase-flip (phase-damping) noise, while has slight effect for suppressing the depolarizing noise. Our study is suitable for JRSP and RSP, and will be helpful for improving the efficiency of multiparticle entanglement based quantum secure communication in real implementation.

Quantum secure communication brings a new direction for information security. As an important component of quantum secure communication, deterministic joint remote state preparation (DJRSP) could securely transmit a quantum state with 100\% success probability. In this paper, we study how the efficiency of DJRSP is affected when qubits involved in the protocol are subjected to noise or decoherence. Taking a GHZ based DJRSP scheme as an example, we study all types of noise usually encountered in real-world implementations of quantum communication protocols, i.e., the bit-flip, phase-flip (phase-damping), depolarizing, and amplitude-damping noise. Our study shows that the fidelity of the output state depends on the phase factor, the amplitude factor and the noise parameter in the bit-flip noise, while the fidelity only depends on the amplitude factor and the noise parameter in the other three types of noise. And the receiver will get different output states depending on the first preparer's measurement result in the amplitude-damping noise. Our results will be helpful for improving quantum secure communication in real implementation.