IEEE Transactions on Fuzzy Systems

Neurofuzzy Networks With Nonlinear Quantum Learning

Massimo Panella, Giovanni Martinelli

Published Mar 9, 2009• 39 citations

Abstract

Nonlinear quantum processing allows the solution of an optimization problem by the exhaustive search on all its possible solutions. Hence, it can replace advantageously the algorithms for learning from a training set. In order to pursue this possibility in the case of neurofuzzy networks, we propose in this paper to tailor their architectures to the requirements of quantum processing. In particular, superposition is introduced to pursue parallelism and entanglement to associate the network performance with each solution present in the superposition. Two aspects of the proposed method are considered in detail: the binary structure of membership functions and fuzzy reasoning and the use of a particular nonlinear quantum algorithm for extracting the optimal neurofuzzy network by exhaustive search.

DOI: 10.1109/tfuzz.2008.928603

Related papers

Neural networks with quantum architecture and quantum learning

Massimo Panella, Giuseppe Martinelli

Abstract A method is proposed for solving the two key problems facing quantum neural networks: introduction of nonlinearity in the neuron operation and efficient use of quantum superposition in the learning algorithm. The former is indirectly solved by using suitable Boolean functions. The latter is based on the use of a suitable nonlinear quantum circuit. The resulting learning procedure does not apply any optimization method. The optimal neural network is obtained by applying an exhaustive search among all the possible solutions. The exhaustive search is carried out by the proposed quantum circuit composed of both linear and nonlinear components. Copyright © 2009 John Wiley & Sons, Ltd.

International Journal of Circuit Theory and ApplicationsJul 23, 200975 citations

Quantum-Inspired Optimization of Neural Network Architecture

Jincheng Zhang

This paper explores a novel approach to neural network architecture optimization inspired by quantum superposition and measurement. Traditional neural network architecture search methods, such as reinforcement learning or evolutionary algorithms, often suffer from computational bottlenecks and slow convergence rates. We propose a genetic algorithm that leverages quantum-inspired probability amplitudes to represent and evaluate neural network architectures. Each architecture is assigned a set of probability amplitudes, reflecting the potential for success. Through a simulated measurement process, the algorithm dynamically adjusts these amplitudes, favoring architectures with higher potential and accelerating the search for optimal configurations. The core idea is that quantum superposition allows us to explore a vast solution space concurrently, while the measurement process mimics the collapse of the wave function, focusing the search on promising candidates. This approach demonstrates the potential to significantly reduce the time and resources required for neural network architecture design, paving the way for more efficient and powerful deep learning models. We present a detailed theoretical framework and outline the key components of the algorithm, emphasizing its ability to mimic quantum phenomena for architectural optimization.

Aug 31, 20260 citations

The Quantum Approach Leading from Evolutionary to Exhaustive Optimization

Massimo Panella, Giuseppe Martinelli

What bio-inspired algorithms mimic are natural mechanisms governing the macroscopic world for optimizing actual performances that are of vital importance. Neural and neurofuzzy networks, genetic, swarm-intelligence and other evolutionary algorithms are well-known results of this imitation. A completely different situation characterizes the microscopic world governed by quantum mechanics. All the possible solutions exist simultaneously in superposition and the problem is to extract the optimal one. In this case, basic mechanisms of quantum mechanics, i.e., superposition and entanglement, are necessary to mimic nature. Following the latter approach, in this paper a quantum architecture was proposed for determining the maximum/minimum in a set of positive integers which is a basic problem related to optimization. The proposed architecture is based on a suitable nonlinear quantum operator and it solves the said problem by an exhaustive search. This was illustrated in detail in the case of a typical NP-complete problem. © 2012 Asian Network for Scientific Information.

Journal of Applied SciencesSep 15, 20122 citations

Artificial neural network architecture selection in a quantum computer

Adenilton José da Silva

Miniaturisation of computers components is taking us from classical to quantum physics domain. Further reduction in computer components size eventually will lead to the development of computer systems whose components will be on such a small scale that quantum physics intrinsic properties must be taken into account. The expression quantum computation and a first formal model of a quantum computer were first employed in the eighties. With the discovery of a quantum algorithm for factoring exponentially faster than any known classical algorithm in 1997, quantum computing began to attract industry investments for the development of a quantum computer and the design of novel quantum algorithms. For instance, the development of learning algorithms for neural networks. Some artificial neural networks models can simulate an universal Turing machine, and together with learning capabilities have numerous applications in real life problems. One limitation of artificial neural networks is the lack of an efficient algorithm to determine its optimal architecture. The main objective of this work is to verify whether we can obtain some advantage with the use of quantum computation techniques in a neural network learning and architecture selection procedure. We propose a quantum neural network, named quantum perceptron over a field (QPF). QPF is a direct generalisation of a classical perceptron which addresses some drawbacks found in previous models for quantum perceptrons. We also present a learning algorithm named Superposition based Architecture Learning algorithm (SAL) that optimises the neural network weights and architectures. SAL searches for the best architecture in a finite set of neural network architectures and neural networks parameters in linear time over the number of examples in the training set. SAL is the first quantum learning algorithm to determine neural network architectures in linear time. This speedup is obtained by the use of quantum parallelism and a non linear quantum operator.

Jun 26, 20150 citations

Superposition Based Learning Algorithm

Adenilton José da Silva, Teresa B. Ludermir, Wilson Rosa de Oliveira

By exploiting the properties of superposition and entanglement found in quantum systems Quantum Computation has been applied to the design of algorithms considerably more efficient than the known classical ones. Known examples are the Shor's factoring algorithm and the Grover's search algorithm. This paper investigates the possibility of employing Quantum Computing techniques to the design of learning algorithms for neural networks tasks such as pattern recognition. We propose a quantum learning algorithm for neural networks where all patterns of the training set are presented concurrently in superposition. In the process we propose a novel model of a quantum weightless neural node. The algorithm is a combination of a quantum search algorithm, a probabilistic quantum memory and a quantum neural network.

Brazilian Symposium on Neural Networks (SBRN)Oct 1, 201010 citations

Quantum HyperNetworks: Training Binary Neural Networks in Quantum Superposition

Juan Carrasquilla, Mohamed Hibat-Allah, Estelle M. Inack +4 more

Binary neural networks, i.e., neural networks whose parameters and activations are constrained to only two possible values, offer a compelling avenue for the deployment of deep learning models on energy- and memory-limited devices. However, their training, architectural design, and hyperparameter tuning remain challenging as these involve multiple computationally expensive combinatorial optimization problems. Here we introduce quantum hypernetworks as a mechanism to train binary neural networks on quantum computers, which unify the search over parameters, hyperparameters, and architectures in a single optimization loop. Through classical simulations, we demonstrate that our approach effectively finds optimal parameters, hyperparameters and architectural choices with high probability on classification problems including a two-dimensional Gaussian dataset and a scaled-down version of the MNIST handwritten digits. We represent our quantum hypernetworks as variational quantum circuits, and find that an optimal circuit depth maximizes the probability of finding performant binary neural networks. Our unified approach provides an immense scope for other applications in the field of machine learning.

arXiv:2301.08292Jan 19, 20231 citation