Quantum neurons with real weights for diabetes prediction

MONTEIRO, Cláudio Luis Alves

Published Sep 2, 2021• 0 citations

Abstract

Parametric models with real numbers valued parameters have greater performance than its counterparts with binary valued weights, due to the gain in representing informa- tion with real values, and therefore having a larger space for memory association. In this work, is proposed a quantum neuron capable of store real weights and preserve the gain of the superposition property, encoding the information in the probability amplitudes of the quantum system, the Real Weights Quantum Neuron. Its performance is compared with other quantum neurons to analyze the application of the quantum neurons on real-world problems, i.e diabetes classification. The results of the experiments shows that a single quantum neuron is capable of achieving an accuracy rate of 100% in the XOR problem and an accuracy rate of 100% in a non-linear dataset, demonstrating that the quantum neurons with real weights are capable of modeling non-linearly separable problems. In the problem of diagnosing diabetes, quantum neurons achieved an accuracy rate of 76% and AUC-ROC of 88%, while its classic version, the perceptron, reached only 63% accuracy and the artificial neural network reached 80% AUC-ROC. These results indicate that a single quantum neuron performs better than its classical version and even the artificial neural network for AUC-ROC, demonstrating potential for use in healthcare applications in the near future. This work is also a contribution to the field of quantum neural networks, which can be further advanced from the quantum neuron proposed.

Related papers

Diabetes Prediction Using Quantum Neurons with Preprocessing Based on Hypercomplex Numbers

Cláudio Alves Monteiro, Fernando M. de Paula Neto

The use of properties that are intrinsic to quantum mechanics has made it possible to build quantum algorithms with greater efficiency than classical algorithms to solve problems whose classically efficient solution either does not exist or is not known. There are quantum neurons that can carry an exponential amount of information to a linear number of quantum information units (qubits) using the quantum property of superposition. In this paper, we compare the performance of three of these quantum neuron models applied to the diabetes classification problem. We also propose the use of different data preprocessing strategies. Quantum neurons were simulated using the IBM Qiskit tool. We compare the preprocessing approaches applied to two toy problems (1) simulating the XOR operator and (2) solving a generic nonlinear problem. The results of the experiments shows that a single quantum neuron is capable of achieving an accuracy rate of 100% in the XOR problem and an accuracy rate of 100% in a non-linear dataset, demonstrating that the quantum neurons with real weights are capable of modeling non-linearly separable problems. In the problem of diagnosing diabetes, quantum neurons achieved an accuracy rate of 76% and AUC-ROC of 88%, while its classic version, the perceptron, reached only 63% accuracy and the artificial neural network reached 80% AUC-ROC. These results indicate that a single quantum neuron performs better than its classical version and even the artificial neural network for AUC-ROC, demonstrating potential for use in healthcare applications in the near future.

IEEE Symposium Series on Computational Intelligence (SSCI)Dec 5, 20211 citation

Quantum Neuron: an elementary building block for machine learning on quantum computers

Yudong Cao, Gian Giacomo Guerreschi, Alán Aspuru‐Guzik

Even the most sophisticated artificial neural networks are built by aggregating substantially identical units called neurons. A neuron receives multiple signals, internally combines them, and applies a non-linear function to the resulting weighted sum. Several attempts to generalize neurons to the quantum regime have been proposed, but all proposals collided with the difficulty of implementing non-linear activation functions, which is essential for classical neurons, due to the linear nature of quantum mechanics. Here we propose a solution to this roadblock in the form of a small quantum circuit that naturally simulates neurons with threshold activation. Our quantum circuit defines a building block, the "quantum neuron", that can reproduce a variety of classical neural network constructions while maintaining the ability to process superpositions of inputs and preserve quantum coherence and entanglement. In the construction of feedforward networks of quantum neurons, we provide numerical evidence that the network not only can learn a function when trained with superposition of inputs and the corresponding output, but that this training suffices to learn the function on all individual inputs separately. When arranged to mimic Hopfield networks, quantum neural networks exhibit properties of associative memory. Patterns are encoded using the simple Hebbian rule for the weights and we demonstrate attractor dynamics from corrupted inputs. Finally, the fact that our quantum model closely captures (traditional) neural network dynamics implies that the vast body of literature and results on neural networks becomes directly relevant in the context of quantum machine learning.

arXiv:1711.11240Nov 30, 2017113 citations

Training an Artificial Neural Network Using Qubits as Artificial Neurons: A Quantum Computing Approach

Avinash Chalumuri, Raghavendra Kune, B. S. B. S. Manoj

Artificial neural networks (ANN) proved to be efficient in solving many problems for big data analytics using machine learning. The complex and non-linear features of the input data can be learned and generalized by ANN. In the big data era, enormous amounts of data arrive from multiple sources. A stage is expected to be reached where even supercomputers are likely inundated with the big data. Training an ANN in such a situation is a challenging task due to the size and dimension of the big data. Also, a large number of parameters are to be used and optimized in the network to learn the patterns and analyze such data. Quantum computing is emerging as a field that provides a solution to this problem as a quantum computer can represent data differently using qubits. Qubits on quantum computers can be used to detect the hidden patterns in data that are difficult for a classical computer to find. Hence, there exists a huge scope for application in the area of artificial neural networks. In this work, we primarily focused on training an artificial neural network using qubits as artificial neurons. The simulation results show that our quantum computing approach for ANN (QC ANN) is efficient when compared to classical ANN. The model with qubits as artificial neurons can learn the features of data using fewer parameters for a binary classification task. We demonstrate our experiment using a quantum simulator and optimization of the quantum parameters used in QC ANN is carried out on a classical computer.

Procedia Computer ScienceJan 1, 202018 citations

Learning Algorithm and Application of Quantum Neural Networks with Quantum Weights

Dianbao Mu, Zunyou Guan, Hong Zhang

A novel neural networks model with quantum weights is present.Firstly, based on the information processing modes of biology neuron and quantum computing theory, a quantum neuron model is presented, which is composed of weighting, aggregating, activating, and inspiriting.Secondly the quantum neural networks model based on quantum neuron is constructed in which the input and the output are real vectors; the linked weight and the activation value are qubits.On the basic of the gradient descent algorithm, a learning algorithm of this model is proposed.It is shown that this algorithm is super-linearly convergent under certain conditions and can increase the probability of getting the global optimal solution.Finally, the availability of the model is illustrated in both convergence speed and convergence rate by two application examples of pattern recognition and function approximation.

International Journal of Computer Theory and EngineeringJan 1, 201327 citations

An artificial neuron implemented on an actual quantum processor

Francesco Tacchino, Chiara Macchiavello, Dario Gerace +1 more

Abstract Artificial neural networks are the heart of machine learning algorithms and artificial intelligence. Historically, the simplest implementation of an artificial neuron traces back to the classical Rosenblatt’s “perceptron”, but its long term practical applications may be hindered by the fast scaling up of computational complexity, especially relevant for the training of multilayered perceptron networks. Here we introduce a quantum information-based algorithm implementing the quantum computer version of a binary-valued perceptron, which shows exponential advantage in storage resources over alternative realizations. We experimentally test a few qubits version of this model on an actual small-scale quantum processor, which gives answers consistent with the expected results. We show that this quantum model of a perceptron can be trained in a hybrid quantum-classical scheme employing a modified version of the perceptron update rule and used as an elementary nonlinear classifier of simple patterns, as a first step towards practical quantum neural networks efficiently implemented on near-term quantum processing hardware.

arXiv:1811.02266npj Quantum InformationMar 29, 2019278 citations

Parametrized constant-depth quantum neuron : framework, conception, and applications

Jonathan Henrique Andrade de CARVALHO

Quantum computing has been revolutionizing the development of algorithms, which in- cludes remarkable advances in artificial neural networks. The exploration of inherently quantum phenomena holds the promise of transcending classical computing. However, only noisy intermediate-scale quantum devices are available currently. To demonstrate advantages in this quantum era, the development of quantum algorithms needs to satisfy several software requirements due to the insufficiency of quantum computing resources. In this research, we propose a kernel-based framework of quantum neurons that not only contemplates existing quantum neurons but also makes room to define countless others, including quantum neurons that comply with the present hardware restrictions. For exam- ple, we propose here a quantum neuron that is implemented by a circuit of constant depth with a linear number of elementary single-qubit gates. Existing quantum neurons are im- plemented by exponentially expensive circuits, even using complex multi-qubit gates. We improve the proposed quantum neuron through a parametrization that can change its ac- tivation function shape in order to fit underlying patterns that existing quantum neurons cannot fit. As an initial demonstration, we show the proposed quantum neuron producing optimal solutions for six classification problems that an existing quantum neuron can solve only two of them. After, we benchmark classical and quantum neurons in several classifi- cation problems. As a result, in the majority of the cases, the proposed quantum neuron is the best over all neurons, which solidly confirms its superiority. The parametrization offers flexibility to not only fit a wide range of problems but also to optimize the margin between classes, at least better than the classical neurons and existing quantum ones. In light of those advantages, this research paves the way to develop quantum neural networks that can demonstrate a practical quantum advantage in the current quantum era already.

Aug 11, 20220 citations