QNet: A Scalable and Noise-Resilient Quantum Neural Network Architecture for Noisy Intermediate-Scale Quantum Computers

Quantum machine learning (QML) is promising for potential speedups and improvements in conventional machine learning (ML) tasks. Existing QML models that use deep parametric quantum circuits (PQC) suffer from a large accumulation of gate errors and decoherence. To circumvent this issue, we propose a new QML architecture called QNet. QNet consists of several small quantum neural networks (QNN). Each of these smaller QNN’s can be executed on small quantum computers that dominate the NISQ-era machines. By carefully choosing the size of these QNN’s, QNet can exploit arbitrary size quantum computers to solve supervised ML tasks of any scale. It also enables heterogeneous technology integration in a single QML application. Through empirical studies, we show the trainability and generalization of QNet and the impact of various configurable variables on its performance. We compare QNet performance against existing models and discuss potential issues and design considerations. In our study, we show 43% better accuracy on average over the existing models on noisy quantum hardware emulators. More importantly, QNet provides a blueprint to build noise-resilient QML models with a collection of small quantum neural networks with near-term noisy quantum devices.

Download Full-text

Federated Quantum Machine Learning

Entropy ◽

10.3390/e23040460 ◽

2021 ◽

Vol 23 (4) ◽

pp. 460

Author(s):

Samuel Yen-Chi Chen ◽

Shinjae Yoo

Keyword(s):

Machine Learning ◽

Data Privacy ◽

Research Direction ◽

Future Research ◽

Quantum Computers ◽

Training Time ◽

Quantum Neural Network ◽

Distributed Training ◽

Machine Learning Model ◽

Quantum Machine Learning

Distributed training across several quantum computers could significantly improve the training time and if we could share the learned model, not the data, it could potentially improve the data privacy as the training would happen where the data is located. One of the potential schemes to achieve this property is the federated learning (FL), which consists of several clients or local nodes learning on their own data and a central node to aggregate the models collected from those local nodes. However, to the best of our knowledge, no work has been done in quantum machine learning (QML) in federation setting yet. In this work, we present the federated training on hybrid quantum-classical machine learning models although our framework could be generalized to pure quantum machine learning model. Specifically, we consider the quantum neural network (QNN) coupled with classical pre-trained convolutional model. Our distributed federated learning scheme demonstrated almost the same level of trained model accuracies and yet significantly faster distributed training. It demonstrates a promising future research direction for scaling and privacy aspects.

Download Full-text

Coreset Clustering on Small Quantum Computers

Electronics ◽

10.3390/electronics10141690 ◽

2021 ◽

Vol 10 (14) ◽

pp. 1690

Author(s):

Teague Tomesh ◽

Pranav Gokhale ◽

Eric R. Anschuetz ◽

Frederic T. Chong

Keyword(s):

Quantum Algorithms ◽

Quantum Computers ◽

Data Sets ◽

New Paradigm ◽

Data Set ◽

Small Quantum ◽

Natural Data ◽

Near Term ◽

Classical Computing ◽

Quantum Speedup

Many quantum algorithms for machine learning require access to classical data in superposition. However, for many natural data sets and algorithms, the overhead required to load the data set in superposition can erase any potential quantum speedup over classical algorithms. Recent work by Harrow introduces a new paradigm in hybrid quantum-classical computing to address this issue, relying on coresets to minimize the data loading overhead of quantum algorithms. We investigated using this paradigm to perform k-means clustering on near-term quantum computers, by casting it as a QAOA optimization instance over a small coreset. We used numerical simulations to compare the performance of this approach to classical k-means clustering. We were able to find data sets with which coresets work well relative to random sampling and where QAOA could potentially outperform standard k-means on a coreset. However, finding data sets where both coresets and QAOA work well—which is necessary for a quantum advantage over k-means on the entire data set—appears to be challenging.

Download Full-text

Machine Learning Methods Applied for Modeling the Process of Obtaining Bricks Using Silicon-Based Materials

Materials ◽

10.3390/ma14237232 ◽

2021 ◽

Vol 14 (23) ◽

pp. 7232

Author(s):

Costel Anton ◽

Silvia Curteanu ◽

Cătălin Lisa ◽

Florin Leon

Keyword(s):

Machine Learning ◽

Neural Networks ◽

Raw Materials ◽

Optimization Procedure ◽

Exhaust Emission ◽

Energy Potential ◽

Learning Methods ◽

Machine Learning Methods ◽

Emission Changes ◽

The Impact

Most of the time, industrial brick manufacture facilities are designed and commissioned for a particular type of manufacture mix and a particular type of burning process. Productivity and product quality maintenance and improvement is a challenge for process engineers. Our paper aims at using machine learning methods to evaluate the impact of adding new auxiliary materials on the amount of exhaust emissions. Experimental determinations made in similar conditions enabled us to build a database containing information about 121 brick batches. Various models (artificial neural networks and regression algorithms) were designed to make predictions about exhaust emission changes when auxiliary materials are introduced into the manufacture mix. The best models were feed-forward neural networks with two hidden layers, having MSE < 0.01 and r2 > 0.82 and, as regression model, kNN with error < 0.6. Also, an optimization procedure, including the best models, was developed in order to determine the optimal values for the parameters that assure the minimum quantities for the gas emission. The Pareto front obtained in the multi-objective optimization conducted with grid search method allows the user the chose the most convenient values for the dry product mass, clay, ash and organic raw materials which minimize gas emissions with energy potential.

Download Full-text

Vaccine nationalism and equitable access to COVID-19 pharmaceuticals: TRIPS Agreement under trial (again)

Journal of International Trade Law and Policy ◽

10.1108/jitlp-03-2021-0012 ◽

2021 ◽

Vol ahead-of-print (ahead-of-print) ◽

Author(s):

Boniface Chimpango

Keyword(s):

Low Income ◽

Empirical Studies ◽

Pharmaceutical Products ◽

Trips Agreement ◽

Low Income Countries ◽

Equitable Access ◽

Content Type ◽

Near Term ◽

Procurement Strategies ◽

The Impact

Purpose The purpose of this study is to contribute towards the debate about global access to COVID-19 vaccines, therapeutics and diagnostics. Design/methodology/approach The global scramble for COVID-19 vaccine and other related pharmaceutical products have once again exposed the limitations of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). High-income countries are claiming a lion’s share of the first available batches of the COVID-19 vaccine in total disregard of the consequences such approach would have on the low-income countries that lack both the manufacturing wherewithal and the financial resources to purchase the vaccine and other products needed to combat the pandemic. This paper reviews the existing TRIPS Flexibilities and analyses their limitations with respect to equitable access of pharmaceutical products in times of health emergencies. This paper then considers the unique challenges that have been brought to the fore by the COVID-19 pandemic. Finally, this paper analytically explores some options that have been proposed so far that the World Trade Organization (WTO) or governments can take in the immediate to near term to facilitate equitable access to COVID-19 pharmaceutical products and technologies. This research is non-empirical, desk-based research. It is, therefore, based on the literature review of existing body of work that is relevant to the topic under discussion. Mindful of the epistemological challenges that are always associated with desk-based research, part of the methodology of this work is to seek support from related empirical studies based on different philosophical underpinnings but that confirm the working hypothesis of this research. Findings This paper finds that there is still a need for a comprehensive reform of TRIPS Agreement to streamline the voluntary licencing system which is an important tool for low-income countries’ access to affordable pharmaceuticals. However, for purposes of dealing with COVID-19, WTO members should consider establishing pooled Licencing Facilities and procurement strategies via already existing political, economic or regional trade groupings. Originality/value This research is original. All sources have been acknowledged. This research synthesises different research papers and applies different viewpoints to the debate on the impact of the TRIPS Agreement on equitable access to COVID-19 vaccines, therapeutics and diagnostics.

Download Full-text

COVID-19 detection on IBM quantum computer with classical-quantum transfer learning

10.1101/2020.11.07.20227306 ◽

2020 ◽

Author(s):

Erdi Acar ◽

İhsan Yilmaz

Keyword(s):

Machine Learning ◽

Transfer Learning ◽

Ct Images ◽

Nucleic Acid Amplification ◽

World Health ◽

Quantum Computers ◽

Data Set ◽

Small Quantum ◽

Short Time

AbstractDiagnose the infected patient as soon as possible in the coronavirus 2019 (COVID-19) outbreak which is declared as a pandemic by the world health organization (WHO) is extremely important. Experts recommend CT imaging as a diagnostic tool because of the weak points of the nucleic acid amplification test (NAAT). In this study, the detection of COVID-19 from CT images, which give the most accurate response in a short time, was investigated in the classical computer and firstly in quantum computers. Using the quantum transfer learning method, we experimentally perform COVID-19 detection in different quantum real processors (IBMQx2, IBMQ-London and IBMQ-Rome) of IBM, as well as in different simulators (Pennylane, Qiskit-Aer and Cirq). By using a small number of data sets such as 126 COVID-19 and 100 Normal CT images, we obtained a positive or negative classification of COVID-19 with 90% success in classical computers, while we achieved a high success rate of 94-100% in quantum computers. Also, according to the results obtained, machine learning process in classical computers requiring more processors and time than quantum computers can be realized in a very short time with a very small quantum processor such as 4 qubits in quantum computers. If the size of the data set is small; Due to the superior properties of quantum, it is seen that according to the classification of COVID-19 and Normal, in terms of machine learning, quantum computers seem to outperform traditional computers.

Download Full-text

Optimizing High-Efficiency Quantum Memory with Quantum Machine Learning for Near-Term Quantum Devices

Scientific Reports ◽

10.1038/s41598-019-56689-0 ◽

2020 ◽

Vol 10 (1) ◽

Cited By ~ 12

Author(s):

Laszlo Gyongyosi ◽

Sandor Imre

Keyword(s):

Machine Learning ◽

Signal To Noise Ratio ◽

Quantum Memory ◽

Quantum Computers ◽

Signal To Noise ◽

Quantum Devices ◽

Retrieval Efficiency ◽

Quantum Machine Learning ◽

Near Term ◽

Quantum Internet

AbstractQuantum memories are a fundamental of any global-scale quantum Internet, high-performance quantum networking and near-term quantum computers. A main problem of quantum memories is the low retrieval efficiency of the quantum systems from the quantum registers of the quantum memory. Here, we define a novel quantum memory called high-retrieval-efficiency (HRE) quantum memory for near-term quantum devices. An HRE quantum memory unit integrates local unitary operations on its hardware level for the optimization of the readout procedure and utilizes the advanced techniques of quantum machine learning. We define the integrated unitary operations of an HRE quantum memory, prove the learning procedure, and evaluate the achievable output signal-to-noise ratio values. We prove that the local unitaries of an HRE quantum memory achieve the optimization of the readout procedure in an unsupervised manner without the use of any labeled data or training sequences. We show that the readout procedure of an HRE quantum memory is realized in a completely blind manner without any information about the input quantum system or about the unknown quantum operation of the quantum register. We evaluate the retrieval efficiency of an HRE quantum memory and the output SNR (signal-to-noise ratio). The results are particularly convenient for gate-model quantum computers and the near-term quantum devices of the quantum Internet.

Download Full-text

ThriftyNets: Convolutional Neural Networks with Tiny Parameter Budget

IoT ◽

10.3390/iot2020012 ◽

2021 ◽

Vol 2 (2) ◽

pp. 222-235

Author(s):

Guillaume Coiffier ◽

Ghouthi Boukli Hacene ◽

Vincent Gripon

Keyword(s):

Neural Network ◽

Machine Learning ◽

Neural Networks ◽

Convolutional Neural Network ◽

Spatial Resolution ◽

Network Architecture ◽

Deep Neural Networks ◽

State Of The Art ◽

Feature Maps ◽

Neural Network Architecture

Deep Neural Networks are state-of-the-art in a large number of challenges in machine learning. However, to reach the best performance they require a huge pool of parameters. Indeed, typical deep convolutional architectures present an increasing number of feature maps as we go deeper in the network, whereas spatial resolution of inputs is decreased through downsampling operations. This means that most of the parameters lay in the final layers, while a large portion of the computations are performed by a small fraction of the total parameters in the first layers. In an effort to use every parameter of a network at its maximum, we propose a new convolutional neural network architecture, called ThriftyNet. In ThriftyNet, only one convolutional layer is defined and used recursively, leading to a maximal parameter factorization. In complement, normalization, non-linearities, downsamplings and shortcut ensure sufficient expressivity of the model. ThriftyNet achieves competitive performance on a tiny parameters budget, exceeding 91% accuracy on CIFAR-10 with less than 40 k parameters in total, 74.3% on CIFAR-100 with less than 600 k parameters, and 67.1% On ImageNet ILSVRC 2012 with no more than 4.15 M parameters. However, the proposed method typically requires more computations than existing counterparts.

Download Full-text

Cross-situational word learning with multimodal neural networks

10.31234/osf.io/udbh2 ◽

2021 ◽

Author(s):

Wai Keen Vong ◽

Brenden M. Lake

Keyword(s):

Neural Network ◽

Machine Learning ◽

Neural Networks ◽

Word Learning ◽

Network Architecture ◽

Question Answering ◽

Cognitive Models ◽

Learning Approaches ◽

Neural Network Learning ◽

Single Epoch

In order to learn the mappings from words to referents, children must integrate co-occurrence information across individually ambiguous pairs of scenes and utterances, a challenge known as cross-situational word learning. In machine learning, recent multimodal neural networks have been shown to learn meaningful visual-linguistic mappings from cross-situational data, as needed to solve problems such as image captioning and visual question answering. These networks are potentially appealing as cognitive models because they can learn from raw visual and linguistic stimuli, something previous cognitive models have not addressed. In this paper, we examine whether recent machine learning approaches can help explain various behavioral phenomena from the psychological literature on cross-situational word learning. We consider two variants of a multimodal neural network architecture, and look at seven different phenomena associated with cross-situational word learning, and word learning more generally. Our results show that these networks can learn word-referent mappings from a single epoch of training, matching the amount of training found in cross-situational word learning experiments. Additionally, these networks capture some, but not all of the phenomena we studied, with all of the failures related to reasoning via mutual exclusivity. These results provide insight into the kinds of phenomena that arise naturally from relatively generic neural network learning algorithms, and which word learning phenomena require additional inductive biases.

Download Full-text

Comparing Class-Aware and Pairwise Loss Functions for Deep Metric Learning in Wildlife Re-Identification

Sensors ◽

10.3390/s21186109 ◽

2021 ◽

Vol 21 (18) ◽

pp. 6109

Author(s):

Nkosikhona Dlamini ◽

Terence L. van Zyl

Keyword(s):

Neural Network ◽

Neural Networks ◽

Loss Function ◽

Network Architecture ◽

Loss Functions ◽

Similarity Learning ◽

Neural Network Architecture ◽

Shot Classification ◽

Public Datasets ◽

The Impact

Similarity learning using deep convolutional neural networks has been applied extensively in solving computer vision problems. This attraction is supported by its success in one-shot and zero-shot classification applications. The advances in similarity learning are essential for smaller datasets or datasets in which few class labels exist per class such as wildlife re-identification. Improving the performance of similarity learning models comes with developing new sampling techniques and designing loss functions better suited to training similarity in neural networks. However, the impact of these advances is tested on larger datasets, with limited attention given to smaller imbalanced datasets such as those found in unique wildlife re-identification. To this end, we test the advances in loss functions for similarity learning on several animal re-identification tasks. We add two new public datasets, Nyala and Lions, to the challenge of animal re-identification. Our results are state of the art on all public datasets tested except Pandas. The achieved Top-1 Recall is 94.8% on the Zebra dataset, 72.3% on the Nyala dataset, 79.7% on the Chimps dataset and, on the Tiger dataset, it is 88.9%. For the Lion dataset, we set a new benchmark at 94.8%. We find that the best performing loss function across all datasets is generally the triplet loss; however, there is only a marginal improvement compared to the performance achieved by Proxy-NCA models. We demonstrate that no single neural network architecture combined with a loss function is best suited for all datasets, although VGG-11 may be the most robust first choice. Our results highlight the need for broader experimentation and exploration of loss functions and neural network architecture for the more challenging task, over classical benchmarks, of wildlife re-identification.

Download Full-text

Statistical Machine Learning in Model Predictive Control of Nonlinear Processes

Mathematics ◽

10.3390/math9161912 ◽

2021 ◽

Vol 9 (16) ◽

pp. 1912

Author(s):

Zhe Wu ◽

David Rincon ◽

Quanquan Gu ◽

Panagiotis D. Christofides

Keyword(s):

Machine Learning ◽

Neural Networks ◽

Model Predictive Control ◽

Predictive Control ◽

Closed Loop ◽

Series Data ◽

Generalization Error ◽

Statistical Machine Learning ◽

The Impact ◽

Loop Stability

Recurrent neural networks (RNNs) have been widely used to model nonlinear dynamic systems using time-series data. While the training error of neural networks can be rendered sufficiently small in many cases, there is a lack of a general framework to guide construction and determine the generalization accuracy of RNN models to be used in model predictive control systems. In this work, we employ statistical machine learning theory to develop a methodological framework of generalization error bounds for RNNs. The RNN models are then utilized to predict state evolution in model predictive controllers (MPC), under which closed-loop stability is established in a probabilistic manner. A nonlinear chemical process example is used to investigate the impact of training sample size, RNN depth, width, and input time length on the generalization error, along with the analyses of probabilistic closed-loop stability through the closed-loop simulations under Lyapunov-based MPC.

Download Full-text