Search Results for author: Songtao Lu

Found 33 papers, 6 papers with code

Improving the Sample and Communication Complexity for Decentralized Non-Convex Optimization: Joint Gradient Estimation and Tracking

no code implementations ICML 2020 Haoran Sun, Songtao Lu, Mingyi Hong

Similarly, for online problems, the proposed method achieves an $\mathcal{O}(m \epsilon^{-3/2})$ sample complexity and an $\mathcal{O}(\epsilon^{-1})$ communication complexity, while the best existing bounds are $\mathcal{O}(m\epsilon^{-2})$ and $\mathcal{O}(\epsilon^{-2})$.

Stochastic Optimization

Min-Max Optimization without Gradients: Convergence and Applications to Black-Box Evasion and Poisoning Attacks

no code implementations ICML 2020 Sijia Liu, Songtao Lu, Xiangyi Chen, Yao Feng, Kaidi Xu, Abdullah Al-Dujaili, Mingyi Hong, Una-May O'Reilly

In this paper, we study the problem of constrained min-max optimization in a black-box setting, where the desired optimizer cannot access the gradients of the objective function but may query its values.

Understanding Benign Overfitting in Nested Meta Learning

no code implementations27 Jun 2022 Lisha Chen, Songtao Lu, Tianyi Chen

While the conventional statistical learning theory suggests that overparameterized models tend to overfit, empirical evidence reveals that overparameterized meta learning methods still work well -- a phenomenon often called ``benign overfitting.''

Few-Shot Learning Learning Theory

Distributed Adversarial Training to Robustify Deep Neural Networks at Scale

1 code implementation13 Jun 2022 Gaoyuan Zhang, Songtao Lu, Yihua Zhang, Xiangyi Chen, Pin-Yu Chen, Quanfu Fan, Lee Martie, Lior Horesh, Mingyi Hong, Sijia Liu

Spurred by that, we propose distributed adversarial training (DAT), a large-batch adversarial training framework implemented over multiple machines.

Distributed Optimization

Min-Max Bilevel Multi-objective Optimization with Applications in Machine Learning

no code implementations3 Mar 2022 Alex Gu, Songtao Lu, Parikshit Ram, Lily Weng

This paper is the first to propose a generic min-max bilevel multi-objective optimization framework, highlighting applications in representation learning and hyperparameter optimization.

Bilevel Optimization Hyperparameter Optimization +3

Taming Communication and Sample Complexities in Decentralized Policy Evaluation for Cooperative Multi-Agent Reinforcement Learning

no code implementations NeurIPS 2021 Xin Zhang, Zhuqing Liu, Jia Liu, Zhengyuan Zhu, Songtao Lu

To our knowledge, this paper is the first work that achieves both $\mathcal{O}(\epsilon^{-2})$ sample complexity and $\mathcal{O}(\epsilon^{-2})$ communication complexity in decentralized policy evaluation for cooperative MARL.

Multi-agent Reinforcement Learning reinforcement-learning +1

Finite-Time Convergence and Sample Complexity of Multi-Agent Actor-Critic Reinforcement Learning with Average Reward

no code implementations ICLR 2022 FNU Hairi, Jia Liu, Songtao Lu

In this paper, we establish the first finite-time convergence result of the actor-critic algorithm for fully decentralized multi-agent reinforcement learning (MARL) problems with average reward.

Multi-agent Reinforcement Learning reinforcement-learning

Understanding Latent Correlation-Based Multiview Learning and Self-Supervision: An Identifiability Perspective

1 code implementation ICLR 2022 Qi Lyu, Xiao Fu, Weiran Wang, Songtao Lu

Under this model, latent correlation maximization is shown to guarantee the extraction of the shared components across views (up to certain ambiguities).

Disentanglement Multiview Learning +1

Signal Transformer: Complex-valued Attention and Meta-Learning for Signal Recognition

no code implementations5 Jun 2021 Yihong Dong, Ying Peng, Muqiao Yang, Songtao Lu, Qingjiang Shi

Deep neural networks have been shown as a class of useful tools for addressing signal recognition issues in recent years, especially for identifying the nonlinear feature structures of signals.

Meta-Learning Time Series

An Efficient Learning Framework For Federated XGBoost Using Secret Sharing And Distributed Optimization

1 code implementation12 May 2021 Lunchen Xie, Jiaqi Liu, Songtao Lu, Tsung-Hui Chang, Qingjiang Shi

XGBoost is one of the most widely used machine learning models in the industry due to its superior learning accuracy and efficiency.

Distributed Optimization

Adversarial Examples can be Effective Data Augmentation for Unsupervised Machine Learning

1 code implementation2 Mar 2021 Chia-Yi Hsu, Pin-Yu Chen, Songtao Lu, Sijia Liu, Chia-Mu Yu

In this paper, we propose a framework of generating adversarial examples for unsupervised models and demonstrate novel applications to data augmentation.

Contrastive Learning Data Augmentation +1

Federated Acoustic Modeling For Automatic Speech Recognition

no code implementations8 Feb 2021 Xiaodong Cui, Songtao Lu, Brian Kingsbury

In this paper, we investigate federated acoustic modeling using data from multiple clients.

Federated Learning Speech Recognition Sound Distributed, Parallel, and Cluster Computing Audio and Speech Processing

Decentralized TD Tracking with Linear Function Approximation and its Finite-Time Analysis

no code implementations NeurIPS 2020 Gang Wang, Songtao Lu, Georgios Giannakis, Gerald Tesauro, Jian Sun

The present contribution deals with decentralized policy evaluation in multi-agent Markov decision processes using temporal-difference (TD) methods with linear function approximation for scalability.

Finding Second-Order Stationary Points Efficiently in Smooth Nonconvex Linearly Constrained Optimization Problems

no code implementations NeurIPS 2020 Songtao Lu, Meisam Razaviyayn, Bo Yang, Kejun Huang, Mingyi Hong

To the best of our knowledge, this is the first time that first-order algorithms with polynomial per-iteration complexity and global sublinear rate are designed to find SOSPs of the important class of non-convex problems with linear constraints (almost surely).

Overcoming Catastrophic Forgetting via Direction-Constrained Optimization

1 code implementation25 Nov 2020 Yunfei Teng, Anna Choromanska, Murray Campbell, Songtao Lu, Parikshit Ram, Lior Horesh

We study the principal directions of the trajectory of the optimizer after convergence and show that traveling along a few top principal directions can quickly bring the parameters outside the cone but this is not the case for the remaining directions.

Continual Learning

Learned Fine-Tuner for Incongruous Few-Shot Adversarial Learning

no code implementations29 Sep 2020 Pu Zhao, Sijia Liu, Parikshit Ram, Songtao Lu, Yuguang Yao, Djallel Bouneffouf, Xue Lin

As novel contributions, we show that the use of LFT within MAML (i) offers the capability to tackle few-shot learning tasks by meta-learning across incongruous yet related problems and (ii) can efficiently work with first-order and derivative-free few-shot learning problems.

Adversarial Attack Few-Shot Learning

Non-convex Min-Max Optimization: Applications, Challenges, and Recent Theoretical Advances

no code implementations15 Jun 2020 Meisam Razaviyayn, Tianjian Huang, Songtao Lu, Maher Nouiehed, Maziar Sanjabi, Mingyi Hong

The min-max optimization problem, also known as the saddle point problem, is a classical optimization problem which is also studied in the context of zero-sum games.

Randomized Bregman Coordinate Descent Methods for Non-Lipschitz Optimization

no code implementations15 Jan 2020 Tianxiang Gao, Songtao Lu, Jia Liu, Chris Chu

Further, we show that the iteration complexity of the proposed method is $O(n\varepsilon^{-2})$ to achieve $\epsilon$-stationary point, where $n$ is the number of blocks of coordinates.

Translation

Distributed Learning in the Non-Convex World: From Batch to Streaming Data, and Beyond

no code implementations14 Jan 2020 Tsung-Hui Chang, Mingyi Hong, Hoi-To Wai, Xinwei Zhang, Songtao Lu

In particular, we {provide a selective review} about the recent techniques developed for optimizing non-convex models (i. e., problem classes), processing batch and streaming data (i. e., data types), over the networks in a distributed manner (i. e., communication and computation paradigm).

Leveraging Two Reference Functions in Block Bregman Proximal Gradient Descent for Non-convex and Non-Lipschitz Problems

no code implementations16 Dec 2019 Tianxiang Gao, Songtao Lu, Jia Liu, Chris Chu

In the applications of signal processing and data analytics, there is a wide class of non-convex problems whose objective function is freed from the common global Lipschitz continuous gradient assumption (e. g., the nonnegative matrix factorization (NMF) problem).

Learn Electronic Health Records by Fully Decentralized Federated Learning

no code implementations4 Dec 2019 Songtao Lu, Yawen Zhang, Yunlong Wang, Christina Mack

Federated learning opens a number of research opportunities due to its high communication efficiency in distributed training problems within a star network.

Federated Learning

No-regret Non-convex Online Meta-Learning

no code implementations22 Oct 2019 Zhenxun Zhuang, Yunlong Wang, Kezi Yu, Songtao Lu

The online meta-learning framework is designed for the continual lifelong learning setting.

Meta-Learning online learning

Improving the Sample and Communication Complexity for Decentralized Non-Convex Optimization: A Joint Gradient Estimation and Tracking Approach

no code implementations13 Oct 2019 Haoran Sun, Songtao Lu, Mingyi Hong

Similarly, for online problems, the proposed method achieves an $\mathcal{O}(m \epsilon^{-3/2})$ sample complexity and an $\mathcal{O}(\epsilon^{-1})$ communication complexity, while the best existing bounds are $\mathcal{O}(m\epsilon^{-2})$ and $\mathcal{O}(\epsilon^{-2})$, respectively.

Stochastic Optimization

Min-Max Optimization without Gradients: Convergence and Applications to Adversarial ML

1 code implementation30 Sep 2019 Sijia Liu, Songtao Lu, Xiangyi Chen, Yao Feng, Kaidi Xu, Abdullah Al-Dujaili, Minyi Hong, Una-May O'Reilly

In this paper, we study the problem of constrained robust (min-max) optimization ina black-box setting, where the desired optimizer cannot access the gradients of the objective function but may query its values.

SNAP: Finding Approximate Second-Order Stationary Solutions Efficiently for Non-convex Linearly Constrained Problems

no code implementations9 Jul 2019 Songtao Lu, Meisam Razaviyayn, Bo Yang, Kejun Huang, Mingyi Hong

This paper proposes low-complexity algorithms for finding approximate second-order stationary points (SOSPs) of problems with smooth non-convex objective and linear constraints.

Understand the dynamics of GANs via Primal-Dual Optimization

no code implementations ICLR 2019 Songtao Lu, Rahul Singh, Xiangyi Chen, Yongxin Chen, Mingyi Hong

By developing new primal-dual optimization tools, we show that, with a proper stepsize choice, the widely used first-order iterative algorithm in training GANs would in fact converge to a stationary solution with a sublinear rate.

Multi-Task Learning

Deep Learning for Signal Demodulation in Physical Layer Wireless Communications: Prototype Platform, Open Dataset, and Analytics

no code implementations8 Mar 2019 Hongmei Wang, Zhenzhen Wu, Shuai Ma, Songtao Lu, Han Zhang, Guoru Ding, Shiyin Li

In this paper, we investigate deep learning (DL)-enabled signal demodulation methods and establish the first open dataset of real modulated signals for wireless communication systems.

Hybrid Block Successive Approximation for One-Sided Non-Convex Min-Max Problems: Algorithms and Applications

no code implementations21 Feb 2019 Songtao Lu, Ioannis Tsaknakis, Mingyi Hong, Yongxin Chen

In this work, we consider a block-wise one-sided non-convex min-max problem, in which the minimization problem consists of multiple blocks and is non-convex, while the maximization problem is (strongly) concave.

Power Market Price Forecasting via Deep Learning

no code implementations18 Sep 2018 Yongli Zhu, Songtao Lu, Renchang Dai, Guangyi Liu, Zhiwei Wang

Then the raw input and output data are preprocessed by unit scaling, and the trained network is tested on the real price data under different input lengths, forecasting horizons and data sizes.

On the Sublinear Convergence of Randomly Perturbed Alternating Gradient Descent to Second Order Stationary Solutions

no code implementations28 Feb 2018 Songtao Lu, Mingyi Hong, Zhengdao Wang

The alternating gradient descent (AGD) is a simple but popular algorithm which has been applied to problems in optimization, machine learning, data ming, and signal processing, etc.

A Nonconvex Splitting Method for Symmetric Nonnegative Matrix Factorization: Convergence Analysis and Optimality

no code implementations24 Mar 2017 Songtao Lu, Mingyi Hong, Zhengdao Wang

The proposed algorithm is guaranteed to converge to the set of Karush-Kuhn-Tucker (KKT) points of the nonconvex SymNMF problem.

Community Detection Semantic Segmentation

Cannot find the paper you are looking for? You can Submit a new open access paper.