arXiv subjects

Tsung-Hui Chang

Publications and source records attributed to Tsung-Hui Chang.

At least 73 records · Page 4Linked to original sources

arXiv2022

Generating Radiology Reports via Memory-driven Transformer

Medical imaging is frequently used in clinical practice and trials for diagnosis and treatment. Writing imaging reports is time-consuming and can be error-prone for inexperienced radiologists. Therefore, automatically generating radiology reports is highly desired to lighten the workload of radiologists and accordingly promote clinical automation, which is an essential task to apply artificial intelligence to the medical domain. In this paper, we propose to generate radiology reports with memory-driven Transformer, where a relational memory is designed to record key information of the generation process and a memory-driven conditional layer normalization is applied to incorporating the memory into the decoder of Transformer. Experimental results on two prevailing radiology report datasets, IU X-Ray and MIMIC-CXR, show that our proposed approach outperforms previous models with respect to both language generation metrics and clinical evaluations. Particularly, this is the first work reporting the generation results on MIMIC-CXR to the best of our knowledge. Further analyses also demonstrate that our approach is able to generate long reports with necessary medical terms as well as meaningful image-text attention mappings.

cs.CL↗

arXiv2022

Federated Stochastic Primal-dual Learning with Differential Privacy

Federated learning (FL) is a new paradigm that enables many clients to jointly train a machine learning (ML) model under the orchestration of a parameter server while keeping the local data not being exposed to any third party. However, the training of FL is an interactive process between local clients and the parameter server. Such process would cause privacy leakage since adversaries may retrieve sensitive information by analyzing the overheard messages. In this paper, we propose a new federated stochastic primal-dual algorithm with differential privacy (FedSPD-DP). Compared to the existing methods, the proposed FedSPD-DP incorporates local stochastic gradient descent (local SGD) and partial client participation (PCP) for addressing the issues of communication efficiency and straggler effects due to randomly accessed clients. Our analysis shows that the data sampling strategy and PCP can enhance the data privacy whereas the larger number of local SGD steps could increase privacy leakage, revealing a non-trivial tradeoff between algorithm communication efficiency and privacy protection. Specifically, we show that, by guaranteeing $(ε, δ)$-DP for each client per communication round, the proposed algorithm guarantees $(\mathcal{O}(qε\sqrt{p T}), δ)$-DP after $T$ communication rounds while maintaining an $\mathcal{O}(1/\sqrt{pTQ})$ convergence rate for a convex and non-smooth learning problem, where $Q$ is the number of local SGD steps, $p$ is the client sampling probability, $q=\max_{i} q_i/\sqrt{1-q_i}$ and $q_i$ is the data sampling probability of each client under PCP. Experiment results are presented to evaluate the practical performance of the proposed algorithm and comparison with state-of-the-art methods.

cs.LG↗

arXiv2022

Quantized Federated Learning under Transmission Delay and Outage Constraints

Federated learning (FL) has been recognized as a viable distributed learning paradigm which trains a machine learning model collaboratively with massive mobile devices in the wireless edge while protecting user privacy. Although various communication schemes have been proposed to expedite the FL process, most of them have assumed ideal wireless channels which provide reliable and lossless communication links between the server and mobile clients. Unfortunately, in practical systems with limited radio resources such as constraint on the training latency and constraints on the transmission power and bandwidth, transmission of a large number of model parameters inevitably suffers from quantization errors (QE) and transmission outage (TO). In this paper, we consider such non-ideal wireless channels, and carry out the first analysis showing that the FL convergence can be severely jeopardized by TO and QE, but intriguingly can be alleviated if the clients have uniform outage probabilities. These insightful results motivate us to propose a robust FL scheme, named FedTOE, which performs joint allocation of wireless resources and quantization bits across the clients to minimize the QE while making the clients have the same TO probability. Extensive experimental results are presented to show the superior performance of FedTOE for deep learning-based classification tasks with transmission latency constraints.

cs.IT↗

arXiv2021

Word Graph Guided Summarization for Radiology Findings

Radiology reports play a critical role in communicating medical findings to physicians. In each report, the impression section summarizes essential radiology findings. In clinical practice, writing impression is highly demanded yet time-consuming and prone to errors for radiologists. Therefore, automatic impression generation has emerged as an attractive research direction to facilitate such clinical practice. Existing studies mainly focused on introducing salient word information to the general text summarization framework to guide the selection of the key content in radiology findings. However, for this task, a model needs not only capture the important words in findings but also accurately describe their relations so as to generate high-quality impressions. In this paper, we propose a novel method for automatic impression generation, where a word graph is constructed from the findings to record the critical words and their relations, then a Word Graph guided Summarization model (WGSum) is designed to generate impressions with the help of the word graph. Experimental results on two datasets, OpenI and MIMIC-CXR, confirm the validity and effectiveness of our proposed approach, where the state-of-the-art results are achieved on both datasets. Further experiments are also conducted to analyze the impact of different graph designs to the performance of our method.

cs.CL↗

arXiv2021

Federated Semi-Supervised Learning with Class Distribution Mismatch

Many existing federated learning (FL) algorithms are designed for supervised learning tasks, assuming that the local data owned by the clients are well labeled. However, in many practical situations, it could be difficult and expensive to acquire complete data labels. Federated semi-supervised learning (Fed-SSL) is an attractive solution for fully utilizing both labeled and unlabeled data. Similar to that encountered in federated supervised learning, class distribution of labeled/unlabeled data could be non-i.i.d. among clients. Besides, in each client, the class distribution of labeled data may be distinct from that of unlabeled data. Unfortunately, both can severely jeopardize the FL performance. To address such challenging issues, we introduce two proper regularization terms that can effectively alleviate the class distribution mismatch problem in Fed-SSL. In addition, to overcome the non-i.i.d. data, we leverage the variance reduction and normalized averaging techniques to develop a novel Fed-SSL algorithm. Theoretically, we prove that the proposed method has a convergence rate of $\mathcal{O}(1/\sqrt{T})$, where $T$ is the number of communication rounds, even when the data distribution are non-i.i.d. among clients. To the best of our knowledge, it is the first formal convergence result for Fed-SSL problems. Numerical experiments based on MNIST data and CIFAR-10 data show that the proposed method can greatly improve the classification accuracy compared to baselines.

cs.LG↗

arXiv2021

Low-rank Matrix Recovery With Unknown Correspondence

We study a matrix recovery problem with unknown correspondence: given the observation matrix $M_o=[A,\tilde P B]$, where $\tilde P$ is an unknown permutation matrix, we aim to recover the underlying matrix $M=[A,B]$. Such problem commonly arises in many applications where heterogeneous data are utilized and the correspondence among them are unknown, e.g., due to privacy concerns. We show that it is possible to recover $M$ via solving a nuclear norm minimization problem under a proper low-rank condition on $M$, with provable non-asymptotic error bound for the recovery of $M$. We propose an algorithm, $\text{M}^3\text{O}$ (Matrix recovery via Min-Max Optimization) which recasts this combinatorial problem as a continuous minimax optimization problem and solves it by proximal gradient with a Max-Oracle. $\text{M}^3\text{O}$ can also be applied to a more general scenario where we have missing entries in $M_o$ and multiple groups of data with distinct unknown correspondence. Experiments on simulated data, the MovieLens 100K dataset and Yale B database show that $\text{M}^3\text{O}$ achieves state-of-the-art performance over several baselines and can recover the ground-truth correspondence with high accuracy.

cs.LG↗

arXiv2021

A Dimension Reduction Technique for Large-scale Structured Sparse Optimization Problems with Application to Convex Clustering

In this paper, we propose a novel adaptive sieving (AS) technique and an enhanced AS (EAS) technique, which are solver independent and could accelerate optimization algorithms for solving large scale convex optimization problems with intrinsic structured sparsity. We establish the finite convergence property of the AS technique and the EAS technique with inexact solutions of the reduced subproblems. As an important application, we apply the AS technique and the EAS technique on the convex clustering model, which could accelerate the state-of-the-art algorithm SSNAL by more than 7 times and the algorithm ADMM by more than 14 times.

math.OC↗

arXiv2021

Clustering by Orthogonal NMF Model and Non-Convex Penalty Optimization

The non-negative matrix factorization (NMF) model with an additional orthogonality constraint on one of the factor matrices, called the orthogonal NMF (ONMF), has been found a promising clustering model and can outperform the classical K-means. However, solving the ONMF model is a challenging optimization problem because the coupling of the orthogonality and non-negativity constraints introduces a mixed combinatorial aspect into the problem due to the determination of the correct status of the variables (positive or zero). Most of the existing methods directly deal with the orthogonality constraint in its original form via various optimization techniques, but are not scalable for large-scale problems. In this paper, we propose a new ONMF based clustering formulation that equivalently transforms the orthogonality constraint into a set of norm-based non-convex equality constraints. We then apply a non-convex penalty (NCP) approach to add them to the objective as penalty terms, leading to a problem that is efficiently solvable. One smooth penalty formulation and one non-smooth penalty formulation are respectively studied. We build theoretical conditions for the penalized problems to provide feasible stationary solutions to the ONMF based clustering problem, as well as proposing efficient algorithms for solving the penalized problems of the two NCP methods. Experimental results based on both synthetic and real datasets are presented to show that the proposed NCP methods are computationally time efficient, and either match or outperform the existing K-means and ONMF based methods in terms of the clustering performance.

cs.LG↗

arXiv2021

Learning to Continuously Optimize Wireless Resource in a Dynamic Environment: A Bilevel Optimization Perspective

There has been a growing interest in developing data-driven, and in particular deep neural network (DNN) based methods for modern communication tasks. For a few popular tasks such as power control, beamforming, and MIMO detection, these methods achieve state-of-the-art performance while requiring less computational efforts, less resources for acquiring channel state information (CSI), etc. However, it is often challenging for these approaches to learn in a dynamic environment. This work develops a new approach that enables data-driven methods to continuously learn and optimize resource allocation strategies in a dynamic environment. Specifically, we consider an ``episodically dynamic" setting where the environment statistics change in ``episodes", and in each episode the environment is stationary. We propose to build the notion of continual learning (CL) into wireless system design, so that the learning model can incrementally adapt to the new episodes, {\it without forgetting} knowledge learned from the previous episodes. Our design is based on a novel bilevel optimization formulation which ensures certain ``fairness" across different data samples. We demonstrate the effectiveness of the CL approach by integrating it with two popular DNN based models for power control and beamforming, respectively, and testing using both synthetic and ray-tracing based data sets. These numerical results show that the proposed CL approach is not only able to adapt to the new scenarios quickly and seamlessly, but importantly, it also maintains high performance over the previously encountered scenarios as well.

eess.SP↗

arXiv2021

Structured Sparse Non-negative Matrix Factorization with L20-Norm for scRNA-seq Data Analysis

Non-negative matrix factorization (NMF) is a powerful tool for dimensionality reduction and clustering. Unfortunately, the interpretation of the clustering results from NMF is difficult, especially for the high-dimensional biological data without effective feature selection. In this paper, we first introduce a row-sparse NMF with $\ell_{2,0}$-norm constraint (NMF_$\ell_{20}$), where the basis matrix $W$ is constrained by the $\ell_{2,0}$-norm, such that $W$ has a row-sparsity pattern with feature selection. It is a challenge to solve the model, because the $\ell_{2,0}$-norm is non-convex and non-smooth. Fortunately, we prove that the $\ell_{2,0}$-norm satisfies the Kurdyka-Ł{ojasiewicz} property. Based on the finding, we present a proximal alternating linearized minimization algorithm and its monotone accelerated version to solve the NMF_$\ell_{20}$ model. In addition, we also present a orthogonal NMF with $\ell_{2,0}$-norm constraint (ONMF_$\ell_{20}$) to enhance the clustering performance by using a non-negative orthogonal constraint. We propose an efficient algorithm to solve ONMF_$\ell_{20}$ by transforming it into a series of constrained and penalized matrix factorization problems. The results on numerical and scRNA-seq datasets demonstrate the efficiency of our methods in comparison with existing methods.

cs.LG↗

arXiv2020

Learning to Continuously Optimize Wireless Resource In Episodically Dynamic Environment

There has been a growing interest in developing data-driven and in particular deep neural network (DNN) based methods for modern communication tasks. For a few popular tasks such as power control, beamforming, and MIMO detection, these methods achieve state-of-the-art performance while requiring less computational efforts, less channel state information (CSI), etc. However, it is often challenging for these approaches to learn in a dynamic environment where parameters such as CSIs keep changing. This work develops a methodology that enables data-driven methods to continuously learn and optimize in a dynamic environment. Specifically, we consider an ``episodically dynamic" setting where the environment changes in ``episodes", and in each episode the environment is stationary. We propose to build the notion of continual learning (CL) into the modeling process of learning wireless systems, so that the learning model can incrementally adapt to the new episodes, {\it without forgetting} knowledge learned from the previous episodes. Our design is based on a novel min-max formulation which ensures certain ``fairness" across different data samples. We demonstrate the effectiveness of the CL approach by customizing it to two popular DNN based models (one for power control and one for beamforming), and testing using both synthetic and real data sets. These numerical results show that the proposed CL approach is not only able to adapt to the new scenarios quickly and seamlessly, but importantly, it maintains high performance over the previously encountered scenarios as well.

eess.SP↗

arXiv2020

Distributed Stochastic Consensus Optimization with Momentum for Nonconvex Nonsmooth Problems

While many distributed optimization algorithms have been proposed for solving smooth or convex problems over the networks, few of them can handle non-convex and non-smooth problems. Based on a proximal primal-dual approach, this paper presents a new (stochastic) distributed algorithm with Nesterov momentum for accelerated optimization of non-convex and non-smooth problems. Theoretically, we show that the proposed algorithm can achieve an $ε$-stationary solution under a constant step size with $\mathcal{O}(1/ε^2)$ computation complexity and $\mathcal{O}(1/ε)$ communication complexity. When compared to the existing gradient tracking based methods, the proposed algorithm has the same order of computation complexity but lower order of communication complexity. To the best of our knowledge, the presented result is the first stochastic algorithm with the $\mathcal{O}(1/ε)$ communication complexity for non-convex and non-smooth problems. Numerical experiments for a distributed non-convex regression problem and a deep neural network based classification problem are presented to illustrate the effectiveness of the proposed algorithms.

math.OC↗

arXiv2020

Learning to Beamform in Heterogeneous Massive MIMO Networks

It is well-known that the problem of finding the optimal beamformers in massive multiple-input multiple-output (MIMO) networks is challenging because of its non-convexity, and conventional optimization based algorithms suffer from high computational costs. While computationally efficient deep learning based methods have been proposed, their complexity heavily relies upon system parameters such as the number of transmit antennas, and therefore these methods typically do not generalize well when deployed in heterogeneous scenarios where the base stations (BSs) are equipped with different numbers of transmit antennas and have different inter-BS distances. This paper proposes a novel deep learning based beamforming algorithm to address the above challenges. Specifically, we consider the weighted sum rate (WSR) maximization problem in multi-input and single-output (MISO) interference channels, and propose a deep neural network architecture by unfolding a parallel gradient projection algorithm. Somewhat surprisingly, by leveraging the low-dimensional structures of the optimal beamforming solution, our constructed neural network can be made independent of the numbers of transmit antennas and BSs. Moreover, such a design can be further extended to a cooperative multicell network. Numerical results based on both synthetic and ray-tracing channel models show that the proposed neural network can achieve high WSRs with significantly reduced runtime, while exhibiting favorable generalization capability with respect to the antenna number, BS number and the inter-BS distance.

cs.IT↗

arXiv2020

Federated Matrix Factorization: Algorithm Design and Application to Data Clustering

Recent demands on data privacy have called for federated learning (FL) as a new distributed learning paradigm in massive and heterogeneous networks. Although many FL algorithms have been proposed, few of them have considered the matrix factorization (MF) model, which is known to have a vast number of signal processing and machine learning applications. Different from the existing FL algorithms that are designed for smooth problems with single block of variables, in federated MF (FedMF), one has to deal with challenging non-convex and non-smooth problems (due to constraints or regularization) with two blocks of variables. In this paper, we address the challenge by proposing two new FedMF algorithms, namely, FedMAvg and FedMGS, based on the model averaging and gradient sharing principles, respectively. Both FedMAvg and FedMGS adopt multiple steps of local updates per communication round to speed up convergence, and allow only a randomly sampled subset of clients to communicate with the server for reducing the communication cost. Convergence analyses for the two algorithms are respectively presented, which delineate the impacts of data distribution, local update number, and partial client communication on the algorithm performance. By focusing on a data clustering task, extensive experiment results are presented to examine the practical performance of both algorithms, as well as demonstrating their efficacy over the existing distributed clustering algorithms.

cs.LG↗

arXiv2020

Learning Structured Communication for Multi-agent Reinforcement Learning

This work explores the large-scale multi-agent communication mechanism under a multi-agent reinforcement learning (MARL) setting. We summarize the general categories of topology for communication structures in MARL literature, which are often manually specified. Then we propose a novel framework termed as Learning Structured Communication (LSC) by using a more flexible and efficient communication topology. Our framework allows for adaptive agent grouping to form different hierarchical formations over episodes, which is generated by an auxiliary task combined with a hierarchical routing protocol. Given each formed topology, a hierarchical graph neural network is learned to enable effective message information generation and propagation among inter- and intra-group communications. In contrast to existing communication mechanisms, our method has an explicit while learnable design for hierarchical communication. Experiments on challenging tasks show the proposed LSC enjoys high communication efficiency, scalability, and global cooperation capability.

cs.LG↗

arXiv2020

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

Distributed learning has become a critical enabler of the massively connected world envisioned by many. This article discusses four key elements of scalable distributed processing and real-time intelligence --- problems, data, communication and computation. Our aim is to provide a fresh and unique perspective about how these elements should work together in an effective and coherent manner. 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). We describe the intuitions and connections behind a core set of popular distributed algorithms, emphasizing how to trade off between computation and communication costs. Practical issues and future research directions will also be discussed.

cs.LG↗

arXiv2019

Transmission Energy Minimization for Heterogeneous Low-Latency NOMA Downlink

This paper investigates the transmission energy minimization problem for the two-user downlink with strictly heterogeneous latency constraints. To cope with the latency constraints and to explicitly specify the trade-off between blocklength (latency) and reliability the normal approximation of the capacity of finite blocklength codes (FBCs) is adopted, in contrast to the classical Shannon capacity formula. We first consider the non-orthogonal multiple access (NOMA) based transmission scheme. However, due to heterogeneous latency constraints and channel conditions at receivers, the conventional successive interference cancellation may be infeasible. We thus study the problem by considering heterogeneous receiver conditions under different interference mitigation schemes and solve the corresponding NOMA design problems. It is shown that, though the energy function is not convex and does not have closed form expression, the studied NOMA problems can be globally solved semi-analytically and with low complexity. Moreover, we propose a hybrid transmission scheme that combines the time division multiple access (TDMA) and NOMA. Specifically, the hybrid scheme can judiciously perform bit and time allocation and take TDMA and NOMA as two special instances. To handle the more challenging hybrid design problem, we propose a concave approximation of the FBC rate/capacity formula, by which we obtain computationally efficient and high-quality solutions. Simulation results show that the hybrid scheme can achieve considerable transmission energy saving compared with both pure NOMA and TDMA schemes.

cs.IT↗

arXiv2019

UAV Positioning and Power Control for Two-Way Wireless Relaying

This paper considers an unmanned-aerial-vehicle-enabled (UAV-enabled) wireless network where a relay UAV is used for two-way communications between a ground base station (BS) and a set of distant user equipment (UE). The UAV adopts the amplify-and-forward strategy for two-way relaying over orthogonal frequency bands. The UAV positioning and the transmission powers of all nodes are jointly designed to maximize the sum rate of both uplink and downlink subject to transmission power constraints and the signal-to-noise ratio constraint on the UAV control channel. The formulated joint positioning and power control (JPPC) problem has an intricate expression of the sum rate due to two-way transmissions and is difficult to solve in general. We propose a novel concave surrogate function for the sum rate and employ the successive convex approximation (SCA) technique for obtaining a high-quality approximate solution. We show that the proposed surrogate function has a small curvature and enables a fast convergence of SCA. Furthermore, we develop a computationally efficient JPPC algorithm by applying the FISTA-type accelerated gradient projection (AGP) algorithm to solve the SCA problem as well as one of the projection subproblem, resulting in a double-loop AGP method. Simulation results show that the proposed JPPC algorithms are not only computationally efficient but also greatly outperform the heuristic approaches.

cs.IT↗