检索结果-内蒙古大学图书馆

2nd mathematical and scientific machine learning conference, msml 2021

The proceedings contain 42 papers. The topics discussed include: deep neural networks are effective at learning high-dimensional Hilbert-valued functions from limited data;temporal-difference learning with nonlinear function approximation: lazy training and mean field regimes;temporal-difference learning with nonlinear function approximation: lazy training and mean field regimes;interpretable and learnable super-resolution time-frequency representation;average-case integrality gap for non-negative principal component analysis;deep autoencoders: from understanding to generalization guarantees;numerical Calabi-Yau metrics from holomorphic networks;some observations on high-dimensional partial differential equations with Barron data;on the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers;reconstruction of pairwise interactions using energy-based models;and ground states of quantum many body lattice models via reinforcement learning.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research on false news detection based on BERT model and attention mechanism 2

Research on false news detection based on BERT model and att...

引用

2nd International conference on Big Data, Computational Intelligence, and Applications, BDCIA 2024

作者： Fang, Jiamin Xi'an Eurasia University Xi'an710000 China

ISBN: (纸本)9781510689053

Artificial intelligence and intelligent algorithms have made remarkable progress in information processing. However, driven by the rapid development of the Internet, fake news interferes with people's access to accurate information. Aiming at the problem of false news detection, a multimodal false news detection model is proposed on the basis of the attention mechanism optimization bidirectional converter model. The designed model has an average accuracy of 98.61% in identifying news and an average response time of 1.74 seconds. It has a high recognition accuracy on various types of news data, especially on plain text news, with an accuracy of 99.82% and a response time of 1.15s. This shows that the new model can accurately and quickly determine the authenticity of news. This study provides an effective fake news identification tool that helps reduce the false information spread and improve the authenticity of network information. © 2025 SPIE.

关键词： Performance modeling Data modeling mathematical optimization Image processing Feature extraction Image enhancement Image fusion Internet machine learning Semantics

来源：评论

学校读者我要写书评

暂无评论

Optimal Policies for a Pandemic: A Stochastic Game Approach and a Deep learning Algorithm 2

Optimal Policies for a Pandemic: A Stochastic Game Approach ...

引用

2nd Annual conference on mathematical and scientific machine learning

作者： Xuan, Yao Balkin, Robert Han, Jiequn Hu, Ruimeng Ceniceros, Hector D. Univ Calif Santa Barbara Dept Math Santa Barbara CA 93106 USA Princeton Univ Dept Math Princeton NJ 08544 USA Univ Calif Santa Barbara Dept Stat & Appl Probabil Santa Barbara CA 93106 USA

Game theory has been an effective tool in the control of disease spread and in suggesting optimal policies at both individual and area levels. In this paper, we propose a multi-region SEIR model based on stochastic differential game theory, aiming to formulate optimal regional policies for infectious diseases. Specifically, we enhance the standard epidemic SEIR model by taking into account the social and health policies issued by multiple region planners. This enhancement makes the model more realistic and powerful. However, it also introduces a formidable computational challenge due to the high dimensionality of the solution space brought by the presence of multiple regions. This significant numerical difficulty of the model structure motivates us to generalize the deep fictitious algorithm introduced in [Han and Hu, msml2020, pp.221-245, PMLR, 2020] and develop an improved algorithm to overcome the curse of dimensionality. We apply the proposed model and algorithm to study the COVID-19 pandemic in three states: New York, New Jersey and Pennsylvania. The model parameters are estimated from real data posted by the Centers for Disease Control and Prevention (CDC). We are able to show the effects of the lockdown/travel ban policy on the spread of COVID-19 for each state and how their policies affect each other.

关键词： Stochastic differential game pandemic optimal policy enhanced deep fictitious play

来源：评论

学校读者我要写书评

暂无评论

machine learning algorithm selection for real-time energy management of hybrid energy ship

引用

ENERGY REPORTS 2022年 8卷 1096-1102页

作者： Gan, Ming Hou, Hui Wu, Xixiu Liu, Bo Yang, Yawei Xie, Changjun Wuhan Univ Technol Sch Automat Wuhan 430070 Peoples R China Wuhan Univ Technol Shenzhen Res Inst Shenzhen 518000 Peoples R China China Ship Dev & Design Ctr Wuhan 430064 Peoples R China

The advantages of hybrid energy ship with photovoltaic in energy conservation and emission reduction are becoming more and more prominent with increasing tension of global fossil energy. However, how to deal with photovoltaic uncertainty in real time and make photovoltaic efficiently connected to ship microgrid has become a key technical problem. Therefore, we propose a real-time energy management strategy for hybrid energy ship based on approximate model predictive control. Firstly, aiming at minimizing the operating cost and deviation from the reference state of charge, an energy management framework based on model predictive control is established. Secondly, the machine learning algorithm is trained to approximate the optimal control action of model predictive control offline, and the performance of different machine learning algorithms is analyzed quantitatively. Finally, taking a ferry equipped with photovoltaic as an example, the appropriate machine learning algorithm and sample number are selected. The results show that the proposed strategy can not only ensure the optimization performance, but also effectively reduce the amount of calculation and realize the real-time operation of energy management in hybrid energy ship. (c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-nd license (http://***/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the 2021 The 2nd International conference on Power Engineering, ICPE, 2021.

关键词： Hybrid energy ship machine learning Approximate model predictive control Real time energy management

来源：评论

学校读者我要写书评

暂无评论

A deep learning method for solving Fokker-Planck equations 2

A deep learning method for solving Fokker-Planck equations

引用

2nd Annual conference on mathematical and scientific machine learning

作者： Zhai, Jiayu Dobson, Matthew Li, Yao Univ Massachusetts Dept Math & Stat Amherst MA 01002 USA

The time evolution of the probability distribution of a stochastic differential equation follows the Fokker-Planck equation, which usually has an unbounded, high-dimensional domain. Inspired by Li (2019), we propose a mesh-free Fokker-Planck solver, in which the solution to the Fokker-Planck equation is now represented by a neural network. The presence of the differential operator in the loss function improves the accuracy of the neural network representation and reduces the demand of data in the training process. Several high dimensional numerical examples are demonstrated.

关键词： Stochastic differential equation Monte Carlo simulation invariant measure coupling method data-driven and machine learning methods

来源：评论

学校读者我要写书评

暂无评论

Simulation of Iterative learning Control Strategy in Powder Metallurgic Forming machine 2

Simulation of Iterative Learning Control Strategy in Powder ...

引用

2nd International conference on Robotics and Automation Sciences (ICRAS)

作者： Gao, Jun Wuhan Business Univ Sch Mech Engn & Automot Serv Wuhan 430056 Hubei Peoples R China

ISBN: (纸本)9781538673706

The principle of the Iterative learning Control (ILC) strategy is introduced in order to research its application in powder metallurgic forming machine. In the hydraulic system of the forming machine, the mathematical model of the main cylinder is introduced in detail. The simulation model is also built based on Matlab/Simulink. The output displacement response of the iterative learning control system is simulated. And, the results show that it is feasible to apply the ILC strategy to hydraulic press machine. The simulation results show that it is feasible to apply the iterative learning control strategy to the position control in the powder metallurgic forming machine.

关键词： powder metallurgic forming position control iterative learning control

来源：评论

学校读者我要写书评

暂无评论

Data-driven power system small-signal stability assessment and correction control model based on XGBoost

引用

ENERGY REPORTS 2022年 8卷 710-717页

作者： Yang, Yude Zhang, Xiu Yang, Lizhen Guangxi Univ Key Lab Power Syst Optimizat & Energy Saving Tech Nanning 530004 Peoples R China Guangxi Univ Sch Elect Engn Nanning 530004 Peoples R China Guangxi Agr Vocat Univ Nanning 530004 Peoples R China

The influence of small-signal stability on the safety and stability of the power system is becoming more prominent. A mapping model based on steady-state operation information is established using the sample learning method, which provides a new technical path for the rapid assessment and correction of significant power grid oscillation characteristics. This paper establishes a small signal stability assessment and correction control model based on the Extreme Gradient Boosting (XGBoost) algorithm. Firstly, the XGBoost model is obtained by analyzing the mapping relationship between generator power, node power, branch power, and minimum damping ratio. Then, the sensitivity of the generator damping ratio is calculated, and the objective is to minimize the active power adjustment amount of the generator. The stability constraint and power balance are the constraint conditions to establish the optimization correction model, obtain the optimal adjustment amount, correct the minimum damping ratio, and improve the system's stability. Finally, the minimum damping ratio after correction is obtained, and the modified damping ratio is estimated by XGBoost algorithm. The performance of the proposed model is verified in IEEE 3-machine 9-node and 10-machine 39-node systems. (c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-nd license (http://***/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the 2021 The 2nd International conference on Power Engineering, ICPE, 2021.

关键词： Small-signal stability assessment Correction XGBoost machine learning

来源：评论

学校读者我要写书评

暂无评论

Ground States of Quantum Many Body Lattice Models via Reinforcement learning 2

Ground States of Quantum Many Body Lattice Models via Reinfo...

引用

2nd Annual conference on mathematical and scientific machine learning

作者： Gispen, Willem Lamacraft, Austen Univ Cambridge Cavendish Lab TCM Grp JJ Thomson Ave Cambridge CB3 0HE England

We introduce reinforcement learning (RL) formulations of the problem of finding the ground state of a many-body quantum mechanical model defined on a lattice. We show that stoquastic Hamiltonians - those without a sign problem - have a natural decomposition into stochastic dynamics and a potential representing a reward function. The mapping to RL is developed for both continuous and discrete time, based on a generalized Feynman-Kac formula in the former case and a stochastic representation of the Schrodinger equation in the latter. We discuss the application of this mapping to the neural representation of quantum states, spelling out the advantages over approaches based on direct representation of the wavefunction of the system.

关键词： Quantum Mechanics Feynman-Kac Formula Optimal Control Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Construction of perfect dispatch learning model based on adaptive GRU

引用

ENERGY REPORTS 2022年 8卷 668-677页

作者： Meng, Xiangfei Zhu, Tanbo Li, Changcheng State Grid Shandong Elect Econ & Technol Res Inst Jinan 250021 Peoples R China State Grid Shandong Elect Power Co Jinan 250021 Peoples R China Guangxi Univ Sch Elect Engn Nanning 530004 Peoples R China

Perfect dispatch refers to finding the after-the-fact optimal allocation of output power among generating units based on actual loads. This paper proposes a new idea to solve the economic dispatch problem that regards it as a learning problem, not an optimal problem. The adaptive Gate Recurrent Unit (GRU) neural network is applied to construct the perfect dispatch learning model, which makes the perfect dispatch from an offline manual analysis method to an automatic online learning method. The hierarchical clustering analysis is performed to determine historical similar days for forecast day. Then the historical days that have the same load characteristic with the forecast day form the training set for the GRU learning model. The matrix correlation analysis is performed to select critical historical instants for each dispatch instant. According to matrix correlation analysis results, the number of the memory blocks in each GRU learning model is determined. The structure of GRU learning models is adaptive for different dispatch instants. The IEEE 39-bus test system is used to verify the feasibility and accuracy of the proposed perfect dispatch learning model. Study results indicate that the generation dispatch schedule obtained by the proposed GRU learning model is closer to the perfect dispatch schedule than the generation dispatch schedule obtained by solving the non-linear optimization problem. (c) 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-nd license (http://***/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of the 2021 The 2nd International conference on Power Engineering, ICPE, 2021.

关键词： Economic dispatch Perfect dispatch machine learning Gate recurrent unit neural network

来源：评论

学校读者我要写书评

暂无评论

Solvable Model for Inheriting the Regularization through Knowledge Distillation 2

Solvable Model for Inheriting the Regularization through Kno...

引用

2nd Annual conference on mathematical and scientific machine learning

作者： Saglietti, Luca Zdeborova, Lenka Ecole Polytech Fed Lausanne SPOC Lab Lausanne Switzerland

In recent years the empirical success of transfer learning with neural networks has stimulated an increasing interest in obtaining a theoretical understanding of its core properties. Knowledge Distillation where a smaller neural network is trained using the outputs of a larger neural network is a particularly interesting case of transfer learning. In the present work, we introduce a statistical physics framework that allows an analytic characterization of the properties of knowledge distillation (KD) in shallow neural networks. Focusing the analysis on a solvable model that exhibits a non-trivial generalization gap, we investigate the effectiveness of KD. We are able to show that, through KD, the regularization properties of the larger teacher model can be inherited by the smaller student and that the yielded generalization performance is closely linked to and limited by the optimality of the teacher. Finally, we analyze the double descent phenomenology that can arise in the considered KD setting.

关键词： Knowledge Distillation Transfer learning Logistic Regression Gaussian Mixture Replica Method

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：