Search

The efficiency of quantum computing has recently been extended to machine learning, which has made a significant impact on quantum machine learning. The hybrid structure of quantum and classical ones has developed into the most successful application mode currently due to noisy intermediate scale quantum limitations. In this paper, an improved hybrid quantum-classic convolutional neural network (HQC-CNN) with fast training speed, lightweight, and high performance is proposed. Its convolution layer realizes feature mapping through parameterized quantum circuit, while other layers keep classic operation and finally complete the task of four classifications of brain tumors. The experiment in this paper is based on kaggle brain tumor magnetic resonance imaging public dataset. The final experimental results show that HQC-CNN can effectively classify meningioma, glioma, pituitary, and no tumor with a classification accuracy of 97.8%. When compared to numerous well-known landmark models, HQC-CNN has obvious advantages. [ABSTRACT FROM AUTHOR]

Shor's factorization algorithm (SFA) aims at finding the non-trivial factor of a given composite number, but the algorithm does not always work. In some cases, it has to call back to the beginning of the algorithm to make recalculation. After the analysis of the principle of SFA and the characteristics of RSA public-key cryptography with a series of data calculations, it can be concluded that the random value selected by the algorithm is closely related to whether the obtained period is effective. Therefore, a new optimized scheme is proposed to tackle with this defect from two perspectives: (a) When the a value is a perfect square, the algorithm can be completed with an odd cycle r. (b) When the period r obtained by the randomly selected a value is a multiple of 3 , the algorithm can be completed by modifying the decomposition method to relax the requirements for the period without affecting the complexity of the algorithm. Due to the limitations of hardware in applying the quantum algorithms, the classical algorithm is applied to simulate quantum algorithms to test the success rate of decomposition of some composite numbers. The result indicates the effectiveness of the improved algorithm, which significantly reduces the probability of repeated operations to save the quantum circuit resources. [ABSTRACT FROM AUTHOR]

Quantum neural networks are neural network models based on the principles of quantum mechanics, a combination of the advantages of classical neural networks and quantum information, using quantum bits instead of neurons in classical neural networks and quantum gates instead of activation functions in classical neural networks. Immune algorithms are a class of algorithms based on the theory of artificial immune systems, which implement functions similar to the biological immune system in terms of antigen recognition, cell differentiation, memory and self-regulation, using a population search strategy and iterative computation to eventually obtain the optimal solution to the problem with a higher probability. Quantum neural networks, one of the most important approaches in the field of quantum machine learning, can be implemented by means of variational quantum circuits. However, the classification training accuracy of variational quantum algorithms is not very high. In this paper, we propose a variational shadow quantum neural network based on the immune optimisation algorithm to classify data in order to improve the training accuracy and also the training speed. Experiments show that the immune optimisation algorithm-based variational shadow quantum neural network can not only improve the training accuracy, but also the training speed. [ABSTRACT FROM AUTHOR]

Previously, single classification models were mainly studied to classify human protein cell images, i.e., to identify a certain protein based on a set of different cells. However, a classifier can identify only one protein, in fact, a single cell usually consists of multiple proteins, and the proteins are not completely independent of each other. In this paper, we build a human protein cell classification model by multi-label learning. The logical relationship and distribution characteristics among the labels are analyzed to determine the different proteins contained in a set of different cells (i.e., containing multiple elements in the output space). In this paper, using human protein image data, we conducted comparison experiments on pre-trained Xception and InceptionResnet V2 to optimize the two models in terms of data augmentation, channel settings, and model structure. The results show that the Optimized InceptionResnet V2 model achieves high performance in the classification task. The final accuracy of the Optimized InceptionResnet V2 model we obtained reached 96.1%, which is a 2.82% improvement relative to that before the optimized model. [ABSTRACT FROM AUTHOR]