搜索结果: 1-15 共查到“数学 learning”相关记录56条 . 查询时间(0.078 秒)
On the Interplay Between Deep Learning and Dynamical Systems
Deep Learning Dynamical Systems 北大
2023/6/16
The explosion of spatiotemporal data in the physical world requires new deep learning tools to model complex dynamical systems.
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Social Learning and Strategic Pricing with Rating Systems
评级系统 社交学习 战略定价
2023/5/15
Despite widespread use in online transactions, rating systems only provide summary statistics of buyers' diverse opinions at best. To investigate the consequences of this coarse form of information ag...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Machine learning methods for forward and inverse PDEs
正反向 偏微分方程 机器学习
2023/4/27
本报告介绍我们近期的两项工作。(1) 求解PDE的PINN方法在处理时间发展方程时往往遇到难以收敛的困难。我们发展了时间方向的预训练PINN方法及自适应步长方法，解决了收敛性困难，使PDE求解精度能够得到系统性提高。我们在一系列时间发展方程上获得了比文献报道更精确的训练结果。(2) 辐射调源问题是一个典型的反问题，要求调整辐射输运方程的边条件(源)，使解满足特定设计目标。我们针对辐射输运方程的时序...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Challenges and Beneficiations of differential privacy in transfer learning and federated learning
迁移学习 联邦学习 差分隐私
2023/4/13
Differential privacy is one of the most common techniques in privacy-preserving machine learning due to its rigorous math definition and theoretical results. Transfer learning and Federated learning a...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Convergence and Implicit Regularization of Deep Learning Optimizers
深度学习 优化 收敛 隐式正则化
2023/4/13
In this talk, I will present our recent work on the convergence/generalization analysis for the popular optimizers in deep learning. (1) We establish the convergence for Adam under (L0,L1 ) smoothness...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Learning Quantum Observables of Low Degrees from a Logarithmic Number of Random Queries
对数 随机查询 低次量子 可观察量
2023/4/14
A fundamental problem from computational learning theory is to well-reconstruct an unknown function on the discrete hypercubes. One classical result of this problem for the random query model is the l...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Real-time tool path planning using deep learning for subtractive manufacturing
深度学习 减材制造 实时刀具 路径规划
2023/4/25
Tool path planning is a crucial factor of computer-aided design and manufacturing (CAD/CAM). To generate suitable tool paths, the previous methods often transform the problem into local or global opti...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Deep learning of multi-scale PDEs based on data generated from particle methods
粒子方法 数据 多尺度 偏微分方程 深度学习
2023/4/26
Solving multi-scale PDEs is difficult in high-dimensional and/or convection-dominant cases. The interacting particle methods (IPM) are shown to outperform solving PDEs directly. Examples include compu...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：Fusion of Supervised Learning and Reinforcement Learning for Dynamic Treatment Recommendation
监督学习 强化学习 融合 动态治疗推荐
2023/5/5
Electronic health records (EHR) have provided a great opportunity to exploit personalized health data to optimize clinical decision making and achieve personalized treatment recommendation. In this ta...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars： Fusion of Supervised Learning and Reinforcement Learning for Dynamic Treatment Recommendation
监督学习 强化学习 融合 动态治疗推荐 电子健康记录
2023/5/6
Electronic health records (EHR) have provided a great opportunity to exploit personalized health data to optimize clinical decision making and achieve personalized treatment recommendation. In this ta...
Academy of Mathematics and Systems Science, CAS Colloquia & Seminars：HT-Net: Hierarchical Transformer based Operator Learning Model for Multiscale PDEs
HT-Net 分层变压器 多尺度 偏微分方程 算子学习模型
2023/5/6
Complex nonlinear interplays of multiple scales give rise to many interesting physical phenomena and pose major difficulties for the computer simulation of multiscale PDE models in areas such as reser...
Last year, MIT researchers announced that they had built “liquid” neural networks, inspired by the brains of small species: a class of flexible, robust machine learning models that learn on the job an...
Tiny machine learning design alleviates a bottleneck in memory usage on internet-of-things devices（图）
林肯实验室 图开发 研讨会 网络科学 虚假信息追踪 机器学习
2023/6/21
As the Covid-19 pandemic has shown, we live in a richly connected world, facilitating not only the efficient spread of a virus but also of information and influence. What can we learn by analyzing the...
2018集成机器学习和预测模拟研讨会（lntegrating Machine Learning and Predictive Simulatio）
2018 集成机器学习和预测模拟 研讨会
2017/12/20
Industrial design and engineering, as well as healthcare, are facing unprecedented challenges with product complexity and increasingly faster innovation cycles. On the one hand, advances in computer m...
In recent years, artificial neural networks a.k.a. deep learning have significantly improved the fields of computer vision, speech recognition, and natural language processing. The success relies on t...