Quantum-inspired Neural Network Research Articles

Quantum-Inspired Neural Network with Runge-Kutta Method

In recent years, researchers have developed novel Quantum-Inspired Neural Network (QINN) frameworks for the Natural Language Processing (NLP) tasks, inspired by the theoretical investigations of quantum cognition. However, we have found that the training efficiency of QINNs is significantly lower than that of classical networks. We analyze the unitary transformation modules of existing QINNs based on the time displacement symmetry of quantum mechanics and discover that they are resembling a mathematical form similar to the first-order Euler method. The high truncation error associated with Euler method affects the training efficiency of QINNs. In order to enhance the training efficiency of QINNs, we generalize QINNs' unitary transformation modules to the Quantum-like high-order Runge-Kutta methods (QRKs). Moreover, we present the results of experiments on conversation emotion recognition and text classification tasks to validate the effectiveness of the proposed approach.

Quantum-Inspired Neural Network Model of Optical Illusions

Ambiguous optical illusions have been a paradigmatic object of fascination, research and inspiration in arts, psychology and video games. However, accurate computational models of perception of ambiguous figures have been elusive. In this paper, we design and train a deep neural network model to simulate human perception of the Necker cube, an ambiguous drawing with several alternating possible interpretations. Defining the weights of the neural network connection using a quantum generator of truly random numbers, in agreement with the emerging concepts of quantum artificial intelligence and quantum cognition, we reveal that the actual perceptual state of the Necker cube is a qubit-like superposition of the two fundamental perceptual states predicted by classical theories. Our results finds applications in video games and virtual reality systems employed for training of astronauts and operators of unmanned aerial vehicles. They are also useful for researchers working in the fields of machine learning and vision, psychology of perception and quantum–mechanical models of human mind and decision making.

Hybrid deep learning and quantum-inspired neural network for day-ahead spatiotemporal wind speed forecasting

Wind is an essential, clean and sustainable renewable source of energy; however, wind speed is stochastic and intermittent. Accurate wind power generation forecasts are required to ensure that power generation can be scheduled economically and securely. This work proposes a hybrid deep learning technique that incorporates a quantum-inspired neural network to predict wind speeds 24 h in advance. An innovative neural network technique is implemented by cascading parallel convolutional neural networks (CNNs) with a long short-term memory (LSTM) and a quantum-inspired neural network (QINN). The proposed hybrid model is optimized using two iterative loops: the outer loop is implemented by using quantum particle swarm optimization (QPSO) to tune the structure of model and other hyperparameters/parameters. The inner loop uses an Adam optimizer to tune the weights and biases of the proposed model. Spatiotemporal wind speeds at various locations provide the 2D input data. Simulation results reveal that the proposed method outperforms other methods for 24 h-ahead wind speed forecasting.

Compressing Color Computer-Generated Hologram Using Gradient Optimized Quantum-Inspired Neural Network

Compressing Color Computer-Generated Hologram Using Gradient Optimized Quantum-Inspired Neural Network

Special Issue on Quantum Inspired Neural Networks for Engineering Optimization

Special Issue on Quantum Inspired Neural Networks for Engineering Optimization

Fast transmission of computer-generated hologram with compressed sensing and quantum-inspired neural network

A two-stage compression framework for computer-generated hologram (CGH) with compressed sensing (CS) and quantum-inspired neural network (QINN) is proposed. A deep learning-based CS model is applied to sample the CGH at sub-Nyquist rates to generate measurements, and its more compact and representative information using interblock information is further extracted by QINN. With two-stage compression framework, the CGH is greatly compressed and suitable for storage and transmission. Experimental results demonstrate that our proposed two-stage compression framework can preserve more important holographic information during the compression, thus significantly improving the overall quality of reconstructed images compared to only using CS or QINN.

Quantum-inspired complex convolutional neural networks

Quantum-inspired neural network is one of the interesting researches at the junction of the two fields of quantum computing and deep learning. Several models of quantum-inspired neurons with real parameters have been proposed, which are mainly used for three-layer feedforward neural networks. In this work, we improve the quantum-inspired neurons by exploiting the complex-valued weights which have richer representational capacity and better non-linearity. We then extend the method of implementing the quantum-inspired neurons to the convolutional operations, and naturally draw the models of quantum-inspired convolutional neural networks (QICNNs) capable of processing high-dimensional data. Five specific structures of QICNNs are discussed which are different in the way of implementing the convolutional and fully connected layers. The performance of classification accuracy of the five QICNNs are tested on the MNIST and CIFAR-10 datasets. The results show that the QICNNs can perform better in classification accuracy on MNIST dataset than the classical CNN. More learning tasks that our QICNN can outperform the classical counterparts will be found.

Novel Authentication and Secure Trust based RPL Routing in Mobile sink supported Internet of Things

ABSTRACT In the modern era, prevalence of the Internet of Things (IoT) devices that have de facto protocol as IPv6 routing protocol for low power and lossy networks (RPL). Yet, RPL protocol is vulnerable to many attacks such as rank attack, password spoofing and more. To this end, most of the works have focused their research on securing the RPL-based IoT network. However, still there exist downsides such as high energy consumption, lack of effective authentication and high packet losses. Motivated by these preceding defects, this paper proposes the Novel Authentication and Secure Trust-based RPL Routing in Mobile sink-supported Internet of Things (SecRPL-MS). At first, SecRPL-MS performs a registration process where all IoT nodes in the network register themselves in the security entity. In this work, the frequent death of IoT nodes is alleviated through deploying mobile sink in the network. If any grid member (GM) node wants to transmit their data to the grid head (GH) node, then it must undergo authentication process. Secure routing is adopted in RPL by utilising the sail fish optimisation algorithm. Each GM node encrypts its sensed data using the prince algorithm before transmitting it to the GH node. The moving points are selected for the mobile sink using the Quantum Inspired Neural Network (QINN) algorithm. This proposed SecRPL-MS performance is evaluated using the Network Simulator 3 (NS3) in terms of the Packet Delivery Ratio (%), Delay (ms), Energy Consumption (mJ), Key Generation Time (ms) and Malicious Node Detection Accuracy (%). The proposed SecRPL-Ms outperforms 23% of malicious node detection accuracy when compared to existing systems, which represent the proposed SecRPL-MS system providing high security by mitigating the following attacks such as rank attack, Sybil attack, blackhole attack and man in the middle attack.

Quantum-inspired Neural Network for Conversational Emotion Recognition

We provide a novel perspective on conversational emotion recognition by drawing an analogy between the task and a complete span of quantum measurement. We characterize different steps of quantum measurement in the process of recognizing speakers' emotions in conversation, and stitch them up with a quantum-like neural network. The quantum-like layers are implemented by complex-valued operations to ensure an authentic adoption of quantum concepts, which naturally enables conversational context modeling and multimodal fusion. We borrow an existing algorithm to learn the complex-valued network weights, so that the quantum-like procedure is conducted in a data-driven manner. Our model is comparable to state-of-the-art approaches on two benchmarking datasets, and provide a quantum view to understand conversational emotion recognition.

Neural Networks for Estimating Speculative Attacks Models.

Currency crises have been analyzed and modeled over the last few decades. These currency crises develop mainly due to a balance of payments crisis, and in many cases, these crises lead to speculative attacks against the price of the currency. Despite the popularity of these models, they are currently shown as models with low estimation precision. In the present study, estimates are made with first- and second-generation speculative attack models using neural network methods. The results conclude that the Quantum-Inspired Neural Network and Deep Neural Decision Trees methodologies are shown to be the most accurate, with results around 90% accuracy. These results exceed the estimates made with Ordinary Least Squares, the usual estimation method for speculative attack models. In addition, the time required for the estimation is less for neural network methods than for Ordinary Least Squares. These results can be of great importance for public and financial institutions when anticipating speculative pressures on currencies that are in price crisis in the markets.

Cortico-Hippocampal Computational Modeling Using Quantum-Inspired Neural Networks.

Many current computational models that aim to simulate cortical and hippocampal modules of the brain depend on artificial neural networks. However, such classical or even deep neural networks are very slow, sometimes taking thousands of trials to obtain the final response with a considerable amount of error. The need for a large number of trials at learning and the inaccurate output responses are due to the complexity of the input cue and the biological processes being simulated. This article proposes a computational model for an intact and a lesioned cortico-hippocampal system using quantum-inspired neural networks. This cortico-hippocampal computational quantum-inspired (CHCQI) model simulates cortical and hippocampal modules by using adaptively updated neural networks entangled with quantum circuits. The proposed model is used to simulate various classical conditioning tasks related to biological processes. The output of the simulated tasks yielded the desired responses quickly and efficiently compared with other computational models, including the recently published Green model.

A Quantum-Inspired Self-Supervised Network model for automatic segmentation of brain MR images

The classical self-supervised neural network architectures suffer from slow convergence problem and incorporation of quantum computing in classical self-supervised networks is a potential solution towards it. In this article, a fully self-supervised novel quantum-inspired neural network model referred to as Quantum-Inspired Self-Supervised Network (QIS-Net) is proposed and tailored for fully automatic segmentation of brain MR images to obviate the challenges faced by deeply supervised Convolutional Neural Network (CNN) architectures. The proposed QIS-Net architecture is composed of three layers of quantum neuron (input, intermediate and output) expressed as qbits. The intermediate and output layers of the QIS-Net architecture are inter-linked through bi-directional propagation of quantum states, wherein the image pixel intensities (quantum bits) are self-organized in between these two layers without any external supervision or training. Quantum observation allows to obtain the true output once the superimposed quantum states interact with the external environment. The proposed self-supervised quantum-inspired network model has been tailored for and tested on Dynamic Susceptibility Contrast (DSC) brain MR images from Nature data sets for detecting complete tumor and reported promising accuracy and reasonable dice similarity scores in comparison with the unsupervised Fuzzy C-Means clustering, self-trained QIBDS Net, Opti-QIBDS Net, deeply supervised U-Net and Fully Convolutional Neural Networks (FCNNs).

Optimized initial weight in quantum-inspired neural network for compressing computer-generated holograms

We propose an optimized initial weight scheme in a quantum-inspired neural network for compressing computer-generated holograms (CGHs). An optimized initial weight generation strategy is applied to accelerate the compression process. The pixel blocks’ complexity distribution of CGH is analyzed, and the parallel quantum neural network structure is used to compress the image pixel blocks. A deep convolutional neural network with residual learning is also adopted for improving the quality of the reconstructed image. The experimental results have shown that the compression iterations are reduced by using the optimized initial weight, and the reconstructed image quality of the compressed CGH is improved using the parallel quantum-inspired neural network structure and the deep convolutional neural network with residual learning.

Optimization of quantum-inspired neural network using memetic algorithm for function approximation and chaotic time series prediction

Abstract Heuristic and deterministic optimization methods are extensively applied for the training of artificial neural networks. Both of these methods have their own advantages and disadvantages. Heuristic stochastic optimization methods like genetic algorithm perform global search, but they suffer from the problem of slow convergence rate near global optimum. On the other hand deterministic methods like gradient descent exhibit a fast convergence rate around global optimum but may get stuck in a local optimum. Motivated by these problems, a hybrid learning algorithm combining genetic algorithm (GA) with gradient descent (GD), called HGAGD, is proposed in this paper. The new algorithm combines the global exploration ability of GA with the accurate local exploitation ability of GD to achieve a faster convergence and also a better accuracy of final solution. The HGAGD is then employed as a new training method to optimize the parameters of a quantum-inspired neural network (QINN) for two different applications. Firstly, two benchmark functions are chosen to demonstrate the potential of the proposed QINN with the HGAGD algorithm in dealing with function approximation problems. Next, the performance of the proposed method in forecasting Mackey–Glass time series and Lorenz attractor is studied. The results of these studies show the superiority of the introduced approach over other published approaches.

Model and Algorithm of Sequence‐Based Quantum‐Inspired Neural Networks

To enhance the approximation ability of traditional Artificial neural network (ANN), by introducing the quantum rotation gates and the multi-qubits controlled-NOT gates to ANN, we proposed a Sequence input-based quantum-inspired neural network (SIQNN). In our model, the hidden nodes are composed of some multi-qubits controlled-NOT gates, the inputs are described by the multi-dimensional discrete qubits sequences, the output nodes are the traditional neurons. The model parameters include the rotation angles of quantum rotation gates in hide layer and the weights in output layer. The learning algorithms were derived by employing the Levenberg-Marquardt algorithm. Simulation results of predicting the runoff of the Hongjiadu Reservoir show that, the SIQNN is obviously superior to the ANN.

Deep quantum inspired neural network with application to aircraft fuel system fault diagnosis

Fault diagnosis for aircraft fuel system can not only improve flight security, but also reduce the huge cost due to regular maintenance. It remains a problem because of the complicated system and the heterogeneous failure modes, especially the different failure modes that have similar impacts on the system. This paper uses the deep quantum inspired neural network (DQINN) which is an improved deep quantum network (DQN) to solve such problem. This method is the combination of classical deep belief network (DBN) and quantum inspired neural network (QINN). For the purpose of inheriting the advantages of DBN and QINN, the structure of DQINN is built in a new fashion. From a system perspective, the DQINN is constructed by the linear superposition of multiple DBNs with quantum intervals in the last hidden layer. Experiments conducted on standard datasets show that DQINN outperforms other three classical algorithms. Finally, a normal model of aircraft fuel system is built and four kinds of common failure modes of the core components are injected into this model, respectively. And the DQINN is applied to the aircraft fuel system fault diagnosis.

Quantum-Inspired Neural Network with Sequence Input

To enhance the approximation and generalization ability of artificial neural network (ANN) by employing the principles of quantum rotation gate and controlled-not gate, a quantum-inspired neuron with sequence input is proposed. In the proposed model, the discrete sequence input is represented by the qubits, which, as the control qubits of the controlled-not gate after being rotated by the quantum rotation gates, control the target qubit for reverse. The model output is described by the probability amplitude of state in the target qubit. Then a quantum-inspired neural network with sequence input (QNNSI) is designed by employing the sequence input-based quantum-inspired neurons to the hidden layer and the classical neurons to the output layer, and a learning algorithm is derived by employing the Levenberg-Marquardt algorithm. Simulation results of benchmark problem show that, under a certain condition, the QNNSI is obviously superior to the ANN.

Quantum-Inspired Neural Network with Quantum Weights and Real Weights

To enhance the approximation ability of neural networks, by introducing quantum rotation gates to the traditional BP networks, a novel quantum-inspired neural network model is proposed in this paper. In our model, the hidden layer consists of quantum neurons. Each quantum neuron carries a group of quantum rotation gates which are used to update the quantum weights. Both input and output layer are composed of the traditional neurons. By employing the back propagation algorithm, the training algorithms are designed. Simulation-based experiments using two application examples of pattern recognition and function approximation, respectively, illustrate the availability of the proposed model.

Quantum-Inspired Neural Networks with Application

In this paper, a novel neural network is proposed based on quantum rotation gate and controlled- NOT gate. Both the input layer and the hide layer are quantum-inspired neurons. The input is given by qubits, and the output is the probability of qubit in the state . By employing the gradient descent method, a training algorithm is introduced. The experimental results show that this model is superior to the common BP networks.

Sequence Input-Based Quantum-Inspired Neural Networks with Applications

To enhance the approximation and generalization ability of artificial neural networks (ANNs) by employing the principle of quantum rotation gate and controlled-not gate, a quantum-inspired neuron with sequence input is proposed. In the proposed model, the discrete sequence input is represented by the qubits, which, as the control qubits of the controlled-not gate after being rotated by the quantum rotation gates, control the target qubit for reverse. The model output is described by the probability amplitude of state $$|1\rangle $$ | 1 ? in the target qubit. Then a quantum-inspired neural networks (QINN) is designed by employing the quantum-inspired neurons to the hidden layer and the common neurons to the output layer. The algorithm of QINN is derived by employing the Levenberg---Marquardt algorithm. Simulation results of some benchmark problems show that, under a certain condition, the QINN is obviously superior to the classical ANN.