Pub Date : 2024-05-12DOI: 10.1088/2632-2153/ad45b1
Shahin Alipour Bonab, Giacomo Russo, Antonio Morandi and Mohammad Yazdani-Asrami
Index-value, or so-called n-value prediction is of paramount importance for understanding the superconductors’ behaviour specially when modeling of superconductors is needed. This parameter is dependent on several physical quantities including temperature, the magnetic field’s density and orientation, and affects the behaviour of high-temperature superconducting devices made out of coated conductors in terms of losses and quench propagation. In this paper, a comprehensive analysis of many machine learning (ML) methods for estimating the n-value has been carried out. The results demonstrated that cascade forward neural network (CFNN) excels in this scope. Despite needing considerably higher training time when compared to the other attempted models, it performs at the highest accuracy, with 0.48 root mean squared error (RMSE) and 99.72% Pearson coefficient for goodness of fit (R-squared). In contrast, the rigid regression method had the worst predictions with 4.92 RMSE and 37.29% R-squared. Also, random forest, boosting methods, and simple feed forward neural network can be considered as a middle accuracy model with faster training time than CFNN. The findings of this study not only advance modeling of superconductors but also pave the way for applications and further research on ML plug-and-play codes for superconducting studies including modeling of superconducting devices.
索引值或所谓的 n 值预测对于理解超导体的行为至关重要，特别是在需要对超导体进行建模时。该参数取决于多个物理量，包括温度、磁场密度和方向，并在损耗和淬火传播方面影响由涂层导体制成的高温超导设备的行为。本文全面分析了用于估算 n 值的多种机器学习（ML）方法。结果表明，级联前向神经网络（CFNN）在这一领域表现出色。尽管与其他尝试过的模型相比，它需要的训练时间要长得多，但它的准确度却最高，均方根误差（RMSE）为 0.48，皮尔逊拟合系数（R-squared）为 99.72%。相比之下，刚性回归法的预测结果最差，均方根误差为 4.92，R 方为 37.29%。此外，随机森林、提升法和简单前馈神经网络可被视为中等精度模型，其训练时间比 CFNN 更快。本研究的发现不仅推进了超导建模，还为超导研究（包括超导设备建模）中的 ML 即插即用代码的应用和进一步研究铺平了道路。
{"title":"A comprehensive machine learning-based investigation for the index-value prediction of 2G HTS coated conductor tapes","authors":"Shahin Alipour Bonab, Giacomo Russo, Antonio Morandi and Mohammad Yazdani-Asrami","doi":"10.1088/2632-2153/ad45b1","DOIUrl":"https://doi.org/10.1088/2632-2153/ad45b1","url":null,"abstract":"Index-value, or so-called n-value prediction is of paramount importance for understanding the superconductors’ behaviour specially when modeling of superconductors is needed. This parameter is dependent on several physical quantities including temperature, the magnetic field’s density and orientation, and affects the behaviour of high-temperature superconducting devices made out of coated conductors in terms of losses and quench propagation. In this paper, a comprehensive analysis of many machine learning (ML) methods for estimating the n-value has been carried out. The results demonstrated that cascade forward neural network (CFNN) excels in this scope. Despite needing considerably higher training time when compared to the other attempted models, it performs at the highest accuracy, with 0.48 root mean squared error (RMSE) and 99.72% Pearson coefficient for goodness of fit (R-squared). In contrast, the rigid regression method had the worst predictions with 4.92 RMSE and 37.29% R-squared. Also, random forest, boosting methods, and simple feed forward neural network can be considered as a middle accuracy model with faster training time than CFNN. The findings of this study not only advance modeling of superconductors but also pave the way for applications and further research on ML plug-and-play codes for superconducting studies including modeling of superconducting devices.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140931911","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-10DOI: 10.1088/2632-2153/ad45af
Javier E Santos, Agnese Marcato, Qinjun Kang, Mohamed Mehana, Daniel O’Malley, Hari Viswanathan, Nicholas Lubbers
Modeling effective transport properties of 3D porous media, such as permeability, at multiple scales is challenging as a result of the combined complexity of the pore structures and fluid physics—in particular, confinement effects which vary across the nanoscale to the microscale. While numerical simulation is possible, the computational cost is prohibitive for realistic domains, which are large and complex. Although machine learning (ML) models have been proposed to circumvent simulation, none so far has simultaneously accounted for heterogeneous 3D structures, fluid confinement effects, and multiple simulation resolutions. By utilizing numerous computer science techniques to improve the scalability of training, we have for the first time developed a general flow model that accounts for the pore-structure and corresponding physical phenomena at scales from Angstrom to the micrometer. Using synthetic computational domains for training, our ML model exhibits strong performance (R2 = 0.9) when tested on extremely diverse real domains at multiple scales.
由于孔隙结构和流体物理--尤其是从纳米尺度到微米尺度各不相同的约束效应--的综合复杂性，对三维多孔介质的有效传输特性（如渗透性）进行多尺度建模极具挑战性。虽然可以进行数值模拟，但对于庞大而复杂的现实领域来说，计算成本过高。虽然已经提出了机器学习（ML）模型来规避模拟，但迄今为止还没有一个模型能同时考虑异质三维结构、流体约束效应和多种模拟分辨率。通过利用大量计算机科学技术来提高训练的可扩展性，我们首次开发了一种通用流动模型，该模型考虑了从埃到微米尺度上的孔隙结构和相应的物理现象。我们的 ML 模型使用合成计算域进行训练，在多种尺度的极其多样化的真实域上进行测试时，表现出很强的性能（R2 = 0.9）。
{"title":"Learning a general model of single phase flow in complex 3D porous media","authors":"Javier E Santos, Agnese Marcato, Qinjun Kang, Mohamed Mehana, Daniel O’Malley, Hari Viswanathan, Nicholas Lubbers","doi":"10.1088/2632-2153/ad45af","DOIUrl":"https://doi.org/10.1088/2632-2153/ad45af","url":null,"abstract":"Modeling effective transport properties of 3D porous media, such as permeability, at multiple scales is challenging as a result of the combined complexity of the pore structures and fluid physics—in particular, confinement effects which vary across the nanoscale to the microscale. While numerical simulation is possible, the computational cost is prohibitive for realistic domains, which are large and complex. Although machine learning (ML) models have been proposed to circumvent simulation, none so far has simultaneously accounted for heterogeneous 3D structures, fluid confinement effects, and multiple simulation resolutions. By utilizing numerous computer science techniques to improve the scalability of training, we have for the first time developed a general flow model that accounts for the pore-structure and corresponding physical phenomena at scales from Angstrom to the micrometer. Using synthetic computational domains for training, our ML model exhibits strong performance (<italic toggle=\"yes\">R</italic>\u0000<sup>2</sup> = 0.9) when tested on extremely diverse real domains at multiple scales.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140931932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-10DOI: 10.1088/2632-2153/ad45b0
Steven Dahdah, James Richard Forbes
This paper proposes a method to identify a Koopman model of a feedback-controlled system given a known controller. The Koopman operator allows a nonlinear system to be rewritten as an infinite-dimensional linear system by viewing it in terms of an infinite set of lifting functions. A finite-dimensional approximation of the Koopman operator can be identified from data by choosing a finite subset of lifting functions and solving a regression problem in the lifted space. Existing methods are designed to identify open-loop systems. However, it is impractical or impossible to run experiments on some systems, such as unstable systems, in an open-loop fashion. The proposed method leverages the linearity of the Koopman operator, along with knowledge of the controller and the structure of the closed-loop (CL) system, to simultaneously identify the CL and plant systems. The advantages of the proposed CL Koopman operator approximation method are demonstrated in simulation using a Duffing oscillator and experimentally using a rotary inverted pendulum system. An open-source software implementation of the proposed method is publicly available, along with the experimental dataset generated for this paper.
{"title":"Closed-loop Koopman operator approximation","authors":"Steven Dahdah, James Richard Forbes","doi":"10.1088/2632-2153/ad45b0","DOIUrl":"https://doi.org/10.1088/2632-2153/ad45b0","url":null,"abstract":"This paper proposes a method to identify a Koopman model of a feedback-controlled system given a known controller. The Koopman operator allows a nonlinear system to be rewritten as an infinite-dimensional linear system by viewing it in terms of an infinite set of lifting functions. A finite-dimensional approximation of the Koopman operator can be identified from data by choosing a finite subset of lifting functions and solving a regression problem in the lifted space. Existing methods are designed to identify open-loop systems. However, it is impractical or impossible to run experiments on some systems, such as unstable systems, in an open-loop fashion. The proposed method leverages the linearity of the Koopman operator, along with knowledge of the controller and the structure of the closed-loop (CL) system, to simultaneously identify the CL and plant systems. The advantages of the proposed CL Koopman operator approximation method are demonstrated in simulation using a Duffing oscillator and experimentally using a rotary inverted pendulum system. An open-source software implementation of the proposed method is publicly available, along with the experimental dataset generated for this paper.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140932006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-09DOI: 10.1088/2632-2153/ad444a
Leonardo Banchi
Many inference scenarios rely on extracting relevant information from known data in order to make future predictions. When the underlying stochastic process satisfies certain assumptions, there is a direct mapping between its exact classical and quantum simulators, with the latter asymptotically using less memory. Here we focus on studying whether such quantum advantage persists when those assumptions are not satisfied, and the model is doomed to have imperfect accuracy. By studying the trade-off between accuracy and memory requirements, we show that quantum models can reach the same accuracy with less memory, or alternatively, better accuracy with the same memory. Finally, we discuss the implications of this result for learning tasks.
{"title":"Accuracy vs memory advantage in the quantum simulation of stochastic processes","authors":"Leonardo Banchi","doi":"10.1088/2632-2153/ad444a","DOIUrl":"https://doi.org/10.1088/2632-2153/ad444a","url":null,"abstract":"Many inference scenarios rely on extracting relevant information from known data in order to make future predictions. When the underlying stochastic process satisfies certain assumptions, there is a direct mapping between its exact classical and quantum simulators, with the latter asymptotically using less memory. Here we focus on studying whether such quantum advantage persists when those assumptions are not satisfied, and the model is doomed to have imperfect accuracy. By studying the trade-off between accuracy and memory requirements, we show that quantum models can reach the same accuracy with less memory, or alternatively, better accuracy with the same memory. Finally, we discuss the implications of this result for learning tasks.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140932078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-09DOI: 10.1088/2632-2153/ad4510
Daniele Soccodato, Gabriele Penazzi, Alessandro Pecchia, Anh-Luan Phan, Matthias Auf der Maur
Empirical tight-binding (ETB) methods have become a common choice to simulate electronic and transport properties for systems composed of thousands of atoms. However, their performance is profoundly dependent on the way the empirical parameters were fitted, and the found parametrizations often exhibit poor transferability. In order to mitigate some of the the criticalities of this method, we introduce a novel Δ-learning scheme, called MLΔTB. After being trained on a custom data set composed of ab-initio band structures, the framework is able to correlate the local atomistic environment to a correction on the on-site ETB parameters, for each atom in the system. The converged algorithm is applied to simulate the electronic properties of random GaAsSb alloys, and displays remarkable agreement both with experimental and ab-initio test data. Some noteworthy characteristics of MLΔTB include the ability to be trained on few instances, to be applied on 3D supercells of arbitrary size, to be rotationally invariant, and to predict physical properties that are not exhibited by the training set.
{"title":"Machine learned environment-dependent corrections for a spds∗ empirical tight-binding basis","authors":"Daniele Soccodato, Gabriele Penazzi, Alessandro Pecchia, Anh-Luan Phan, Matthias Auf der Maur","doi":"10.1088/2632-2153/ad4510","DOIUrl":"https://doi.org/10.1088/2632-2153/ad4510","url":null,"abstract":"Empirical tight-binding (ETB) methods have become a common choice to simulate electronic and transport properties for systems composed of thousands of atoms. However, their performance is profoundly dependent on the way the empirical parameters were fitted, and the found parametrizations often exhibit poor transferability. In order to mitigate some of the the criticalities of this method, we introduce a novel Δ-learning scheme, called MLΔTB. After being trained on a custom data set composed of <italic toggle=\"yes\">ab-initio</italic> band structures, the framework is able to correlate the local atomistic environment to a correction on the on-site ETB parameters, for each atom in the system. The converged algorithm is applied to simulate the electronic properties of random GaAsSb alloys, and displays remarkable agreement both with experimental and <italic toggle=\"yes\">ab-initio</italic> test data. Some noteworthy characteristics of MLΔTB include the ability to be trained on few instances, to be applied on 3D supercells of arbitrary size, to be rotationally invariant, and to predict physical properties that are not exhibited by the training set.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140931909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-09DOI: 10.1088/2632-2153/ad450f
Antonio Ferrer-Sánchez, Carlos Flores-Garrigos, Carlos Hernani-Morales, José J Orquín-Marqués, Narendra N Hegade, Alejandro Gomez Cadavid, Iraitz Montalban, Enrique Solano, Yolanda Vives-Gilabert, José D Martín-Guerrero
A novel methodology that leverages physics-informed neural networks to optimize quantum circuits in systems with NQ qubits by addressing the counterdiabatic (CD) protocol is introduced. The primary purpose is to employ physics-inspired deep learning techniques for accurately modeling the time evolution of various physical observables within quantum systems. To achieve this, we integrate essential physical information into an underlying neural network to effectively tackle the problem. Specifically, the imposition of the solution to meet the principle of least action, along with the hermiticity condition on all physical observables, among others, ensuring the acquisition of appropriate CD terms based on underlying physics. This approach provides a reliable alternative to previous methodologies relying on classical numerical approximations, eliminating their inherent constraints. The proposed method offers a versatile framework for optimizing physical observables relevant to the problem, such as the scheduling function, gauge potential, temporal evolution of energy levels, among others. This methodology has been successfully applied to 2-qubit representing H2 molecule using the STO-3G basis, demonstrating the derivation of a desirable decomposition for non-adiabatic terms through a linear combination of Pauli operators. This attribute confers significant advantages for practical implementation within quantum computing algorithms.
{"title":"Physics-informed neural networks for an optimal counterdiabatic quantum computation","authors":"Antonio Ferrer-Sánchez, Carlos Flores-Garrigos, Carlos Hernani-Morales, José J Orquín-Marqués, Narendra N Hegade, Alejandro Gomez Cadavid, Iraitz Montalban, Enrique Solano, Yolanda Vives-Gilabert, José D Martín-Guerrero","doi":"10.1088/2632-2153/ad450f","DOIUrl":"https://doi.org/10.1088/2632-2153/ad450f","url":null,"abstract":"A novel methodology that leverages physics-informed neural networks to optimize quantum circuits in systems with <inline-formula>\u0000<tex-math><?CDATA $mathrm{N}_{mathrm{Q}}$?></tex-math>\u0000<mml:math overflow=\"scroll\"><mml:mrow><mml:msub><mml:mrow><mml:mi mathvariant=\"normal\">N</mml:mi></mml:mrow><mml:mrow><mml:mrow><mml:mi mathvariant=\"normal\">Q</mml:mi></mml:mrow></mml:mrow></mml:msub></mml:mrow></mml:math>\u0000<inline-graphic xlink:href=\"mlstad450fieqn1.gif\" xlink:type=\"simple\"></inline-graphic>\u0000</inline-formula> qubits by addressing the counterdiabatic (CD) protocol is introduced. The primary purpose is to employ physics-inspired deep learning techniques for accurately modeling the time evolution of various physical observables within quantum systems. To achieve this, we integrate essential physical information into an underlying neural network to effectively tackle the problem. Specifically, the imposition of the solution to meet the principle of least action, along with the hermiticity condition on all physical observables, among others, ensuring the acquisition of appropriate CD terms based on underlying physics. This approach provides a reliable alternative to previous methodologies relying on classical numerical approximations, eliminating their inherent constraints. The proposed method offers a versatile framework for optimizing physical observables relevant to the problem, such as the scheduling function, gauge potential, temporal evolution of energy levels, among others. This methodology has been successfully applied to 2-qubit representing <inline-formula>\u0000<tex-math><?CDATA $mathrm{H}_{2}$?></tex-math>\u0000<mml:math overflow=\"scroll\"><mml:mrow><mml:msub><mml:mrow><mml:mi mathvariant=\"normal\">H</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msub></mml:mrow></mml:math>\u0000<inline-graphic xlink:href=\"mlstad450fieqn2.gif\" xlink:type=\"simple\"></inline-graphic>\u0000</inline-formula> molecule using the STO-3G basis, demonstrating the derivation of a desirable decomposition for non-adiabatic terms through a linear combination of Pauli operators. This attribute confers significant advantages for practical implementation within quantum computing algorithms.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140931908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-06DOI: 10.1088/2632-2153/ad4230
Ihda Chaerony Siffa, Markus M Becker, Klaus-Dieter Weltmann and Jan Trieschmann
Poisson’s equation plays an important role in modeling many physical systems. In electrostatic self-consistent low-temperature plasma (LTP) simulations, Poisson’s equation is solved at each simulation time step, which can amount to a significant computational cost for the entire simulation. In this paper, we describe the development of a generic machine-learned Poisson solver specifically designed for the requirements of LTP simulations in complex 2D reactor geometries on structured Cartesian grids. Here, the reactor geometries can consist of inner electrodes and dielectric materials as often found in LTP simulations. The approach leverages a hybrid CNN-transformer network architecture in combination with a weighted multiterm loss function. We train the network using highly randomized synthetic data to ensure the generalizability of the learned solver to unseen reactor geometries. The results demonstrate that the learned solver is able to produce quantitatively and qualitatively accurate solutions. Furthermore, it generalizes well on new reactor geometries such as reference geometries found in the literature. To increase the numerical accuracy of the solutions required in LTP simulations, we employ a conventional iterative solver to refine the raw predictions, especially to recover the high-frequency features not resolved by the initial prediction. With this, the proposed learned Poisson solver provides the required accuracy and is potentially faster than a pure GPU-based conventional iterative solver. This opens up new possibilities for developing a generic and high-performing learned Poisson solver for LTP systems in complex geometries.
{"title":"Towards a machine-learned Poisson solver for low-temperature plasma simulations in complex geometries","authors":"Ihda Chaerony Siffa, Markus M Becker, Klaus-Dieter Weltmann and Jan Trieschmann","doi":"10.1088/2632-2153/ad4230","DOIUrl":"https://doi.org/10.1088/2632-2153/ad4230","url":null,"abstract":"Poisson’s equation plays an important role in modeling many physical systems. In electrostatic self-consistent low-temperature plasma (LTP) simulations, Poisson’s equation is solved at each simulation time step, which can amount to a significant computational cost for the entire simulation. In this paper, we describe the development of a generic machine-learned Poisson solver specifically designed for the requirements of LTP simulations in complex 2D reactor geometries on structured Cartesian grids. Here, the reactor geometries can consist of inner electrodes and dielectric materials as often found in LTP simulations. The approach leverages a hybrid CNN-transformer network architecture in combination with a weighted multiterm loss function. We train the network using highly randomized synthetic data to ensure the generalizability of the learned solver to unseen reactor geometries. The results demonstrate that the learned solver is able to produce quantitatively and qualitatively accurate solutions. Furthermore, it generalizes well on new reactor geometries such as reference geometries found in the literature. To increase the numerical accuracy of the solutions required in LTP simulations, we employ a conventional iterative solver to refine the raw predictions, especially to recover the high-frequency features not resolved by the initial prediction. With this, the proposed learned Poisson solver provides the required accuracy and is potentially faster than a pure GPU-based conventional iterative solver. This opens up new possibilities for developing a generic and high-performing learned Poisson solver for LTP systems in complex geometries.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140888473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-06DOI: 10.1088/2632-2153/ad43b1
A Bal, T Brandes, F Iemmi, M Klute, B Maier, V Mikuni and T K Årrestad
Knowledge distillation is a form of model compression that allows artificial neural networks of different sizes to learn from one another. Its main application is the compactification of large deep neural networks to free up computational resources, in particular on edge devices. In this article, we consider proton-proton collisions at the High-Luminosity Large Hadron Collider (HL-LHC) and demonstrate a successful knowledge transfer from an event-level graph neural network (GNN) to a particle-level small deep neural network (DNN). Our algorithm, DistillNet, is a DNN that is trained to learn about the provenance of particles, as provided by the soft labels that are the GNN outputs, to predict whether or not a particle originates from the primary interaction vertex. The results indicate that for this problem, which is one of the main challenges at the HL-LHC, there is minimal loss during the transfer of knowledge to the small student network, while improving significantly the computational resource needs compared to the teacher. This is demonstrated for the distilled student network on a CPU, as well as for a quantized and pruned student network deployed on an field programmable gate array. Our study proves that knowledge transfer between networks of different complexity can be used for fast artificial intelligence (AI) in high-energy physics that improves the expressiveness of observables over non-AI-based reconstruction algorithms. Such an approach can become essential at the HL-LHC experiments, e.g. to comply with the resource budget of their trigger stages.
{"title":"Distilling particle knowledge for fast reconstruction at high-energy physics experiments","authors":"A Bal, T Brandes, F Iemmi, M Klute, B Maier, V Mikuni and T K Årrestad","doi":"10.1088/2632-2153/ad43b1","DOIUrl":"https://doi.org/10.1088/2632-2153/ad43b1","url":null,"abstract":"Knowledge distillation is a form of model compression that allows artificial neural networks of different sizes to learn from one another. Its main application is the compactification of large deep neural networks to free up computational resources, in particular on edge devices. In this article, we consider proton-proton collisions at the High-Luminosity Large Hadron Collider (HL-LHC) and demonstrate a successful knowledge transfer from an event-level graph neural network (GNN) to a particle-level small deep neural network (DNN). Our algorithm, DistillNet, is a DNN that is trained to learn about the provenance of particles, as provided by the soft labels that are the GNN outputs, to predict whether or not a particle originates from the primary interaction vertex. The results indicate that for this problem, which is one of the main challenges at the HL-LHC, there is minimal loss during the transfer of knowledge to the small student network, while improving significantly the computational resource needs compared to the teacher. This is demonstrated for the distilled student network on a CPU, as well as for a quantized and pruned student network deployed on an field programmable gate array. Our study proves that knowledge transfer between networks of different complexity can be used for fast artificial intelligence (AI) in high-energy physics that improves the expressiveness of observables over non-AI-based reconstruction algorithms. Such an approach can become essential at the HL-LHC experiments, e.g. to comply with the resource budget of their trigger stages.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140888488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-06DOI: 10.1088/2632-2153/ad43b3
Carlos Granero Belinchon and Manuel Cabeza Gallucci
This article introduces a new neural network stochastic model to generate a 1-dimensional stochastic field with turbulent velocity statistics. Both the model architecture and training procedure ground on the Kolmogorov and Obukhov statistical theories of fully developed turbulence, so guaranteeing descriptions of (1) energy distribution, (2) energy cascade and (3) intermittency across scales in agreement with experimental observations. The model is a generative adversarial network (GAN) with multiple multiscale optimization criteria. First, we use three physics-based criteria: the variance, skewness and flatness of the increments of the generated field, that retrieve respectively the turbulent energy distribution, energy cascade and intermittency across scales. Second, the GAN criterion, based on reproducing statistical distributions, is used on segments of different length of the generated field. Furthermore, to mimic multiscale decompositions frequently used in turbulence’s studies, the model architecture is fully convolutional with kernel sizes varying along the multiple layers of the model. To train our model, we use turbulent velocity signals from grid turbulence at Modane wind tunnel.
本文介绍了一种新的神经网络随机模型，用于生成具有湍流速度统计量的一维随机场。该模型的结构和训练程序都基于湍流充分发展的科尔莫哥洛夫和奥布霍夫统计理论，因此能保证与实验观测结果一致地描述（1）能量分布、（2）能量级联和（3）跨尺度的间歇性。该模型是一个具有多种多尺度优化标准的生成式对抗网络（GAN）。首先，我们使用三个基于物理学的标准：生成场增量的方差、偏斜度和平坦度，它们分别检索湍流能量分布、能量级联和跨尺度间歇性。其次，基于再现统计分布的 GAN 标准被用于生成场的不同长度段。此外，为了模仿湍流研究中常用的多尺度分解，模型架构采用了全卷积方式，核大小随模型的多个层而变化。为了训练我们的模型，我们使用了来自莫达纳风洞网格湍流的湍流速度信号。
{"title":"A multiscale and multicriteria generative adversarial network to synthesize 1-dimensional turbulent fields","authors":"Carlos Granero Belinchon and Manuel Cabeza Gallucci","doi":"10.1088/2632-2153/ad43b3","DOIUrl":"https://doi.org/10.1088/2632-2153/ad43b3","url":null,"abstract":"This article introduces a new neural network stochastic model to generate a 1-dimensional stochastic field with turbulent velocity statistics. Both the model architecture and training procedure ground on the Kolmogorov and Obukhov statistical theories of fully developed turbulence, so guaranteeing descriptions of (1) energy distribution, (2) energy cascade and (3) intermittency across scales in agreement with experimental observations. The model is a generative adversarial network (GAN) with multiple multiscale optimization criteria. First, we use three physics-based criteria: the variance, skewness and flatness of the increments of the generated field, that retrieve respectively the turbulent energy distribution, energy cascade and intermittency across scales. Second, the GAN criterion, based on reproducing statistical distributions, is used on segments of different length of the generated field. Furthermore, to mimic multiscale decompositions frequently used in turbulence’s studies, the model architecture is fully convolutional with kernel sizes varying along the multiple layers of the model. To train our model, we use turbulent velocity signals from grid turbulence at Modane wind tunnel.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140881389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-05-02DOI: 10.1088/2632-2153/ad3ee6
Jaime Carracedo-Cosme, Prokop Hapala and Rubén Pérez
Atomic force microscopy (AFM) operating in the frequency modulation mode with a metal tip functionalized with a CO molecule is able to image the internal structure of molecules with an unprecedented resolution. The interpretation of these images is often difficult, making the support of theoretical simulations important. Current simulation methods, particularly the most accurate ones, require expertise and resources to perform ab initio calculations for the necessary inputs (i.e charge density and electrostatic potential of the molecule). Here, we propose a computationally inexpensive and fast alternative to the physical simulation of these AFM images based on a conditional generative adversarial network (CGAN), that avoids all force calculations, and uses as the only input a 2D ball–and–stick depiction of the molecule. We discuss the performance of the model when trained with different subsets extracted from the previously published QUAM-AFM database. Our CGAN reproduces accurately the intramolecular contrast observed in the simulated images for quasi–planar molecules, but has limitations for molecules with a substantial internal corrugation, due to the strictly 2D character of the input.
原子力显微镜（AFM）在频率调制模式下工作，其金属尖端具有 CO 分子功能，能够以前所未有的分辨率对分子的内部结构进行成像。对这些图像的解读通常比较困难，因此理论模拟的支持就显得尤为重要。目前的模拟方法，尤其是最精确的模拟方法，需要专业知识和资源来对必要的输入（即分子的电荷密度和静电势）进行 ab initio 计算。在此，我们提出了一种基于条件生成对抗网络（CGAN）的计算成本低廉、速度快的替代方法，它避免了所有的力计算，仅使用分子的二维球棍描绘作为唯一输入，对这些原子力显微镜图像进行物理模拟。我们讨论了该模型在使用从之前发布的 QUAM-AFM 数据库中提取的不同子集进行训练时的性能。对于准平面分子，我们的 CGAN 能准确再现模拟图像中观察到的分子内对比度，但对于具有大量内部波纹的分子，由于输入的严格二维特征，我们的 CGAN 有其局限性。
{"title":"Atomic force microscopy simulations for CO-functionalized tips with deep learning","authors":"Jaime Carracedo-Cosme, Prokop Hapala and Rubén Pérez","doi":"10.1088/2632-2153/ad3ee6","DOIUrl":"https://doi.org/10.1088/2632-2153/ad3ee6","url":null,"abstract":"Atomic force microscopy (AFM) operating in the frequency modulation mode with a metal tip functionalized with a CO molecule is able to image the internal structure of molecules with an unprecedented resolution. The interpretation of these images is often difficult, making the support of theoretical simulations important. Current simulation methods, particularly the most accurate ones, require expertise and resources to perform ab initio calculations for the necessary inputs (i.e charge density and electrostatic potential of the molecule). Here, we propose a computationally inexpensive and fast alternative to the physical simulation of these AFM images based on a conditional generative adversarial network (CGAN), that avoids all force calculations, and uses as the only input a 2D ball–and–stick depiction of the molecule. We discuss the performance of the model when trained with different subsets extracted from the previously published QUAM-AFM database. Our CGAN reproduces accurately the intramolecular contrast observed in the simulated images for quasi–planar molecules, but has limitations for molecules with a substantial internal corrugation, due to the strictly 2D character of the input.","PeriodicalId":33757,"journal":{"name":"Machine Learning Science and Technology","volume":null,"pages":null},"PeriodicalIF":6.8,"publicationDate":"2024-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140840481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：