arXiv subjects

Jiayu Zhou

Publications and source records attributed to Jiayu Zhou.

At least 73 records · Page 4Linked to original sources

arXiv2022

Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization

Federated learning (FL) provides a distributed learning framework for multiple participants to collaborate learning without sharing raw data. In many practical FL scenarios, participants have heterogeneous resources due to disparities in hardware and inference dynamics that require quickly loading models of different sizes and levels of robustness. The heterogeneity and dynamics together impose significant challenges to existing FL approaches and thus greatly limit FL's applicability. In this paper, we propose a novel Split-Mix FL strategy for heterogeneous participants that, once training is done, provides in-situ customization of model sizes and robustness. Specifically, we achieve customization by learning a set of base sub-networks of different sizes and robustness levels, which are later aggregated on-demand according to inference requirements. This split-mix strategy achieves customization with high efficiency in communication, storage, and inference. Extensive experiments demonstrate that our method provides better in-situ customization than the existing heterogeneous-architecture FL methods. Codes and pre-trained models are available: https://github.com/illidanlab/SplitMix.

cs.LG↗

arXiv2022

Margin Preserving Self-paced Contrastive Learning Towards Domain Adaptation for Medical Image Segmentation

To bridge the gap between the source and target domains in unsupervised domain adaptation (UDA), the most common strategy puts focus on matching the marginal distributions in the feature space through adversarial learning. However, such category-agnostic global alignment lacks of exploiting the class-level joint distributions, causing the aligned distribution less discriminative. To address this issue, we propose in this paper a novel margin preserving self-paced contrastive Learning (MPSCL) model for cross-modal medical image segmentation. Unlike the conventional construction of contrastive pairs in contrastive learning, the domain-adaptive category prototypes are utilized to constitute the positive and negative sample pairs. With the guidance of progressively refined semantic prototypes, a novel margin preserving contrastive loss is proposed to boost the discriminability of embedded representation space. To enhance the supervision for contrastive learning, more informative pseudo-labels are generated in target domain in a self-paced way, thus benefiting the category-aware distribution alignment for UDA. Furthermore, the domain-invariant representations are learned through joint contrastive learning between the two domains. Extensive experiments on cross-modal cardiac segmentation tasks demonstrate that MPSCL significantly improves semantic segmentation performance, and outperforms a wide variety of state-of-the-art methods by a large margin.

cs.CV↗

arXiv2021

Persona Authentication through Generative Dialogue

In this paper we define and investigate the problem of \emph{persona authentication}: learning a conversational policy to verify the consistency of persona models. We propose a learning objective and prove (under some mild assumptions) that local density estimators trained under this objective maximize the mutual information between persona information and dialog trajectory. Based on the proposed objective, we develop a method of learning an authentication model that adaptively outputs personalized questions to reveal the underlying persona of its partner throughout the course of multi-turn conversation. Experiments show that our authentication method discovers effective question sequences that generalize to unseen persona profiles.

cs.CL↗

arXiv2021

FedFace: Collaborative Learning of Face Recognition Model

DNN-based face recognition models require large centrally aggregated face datasets for training. However, due to the growing data privacy concerns and legal restrictions, accessing and sharing face datasets has become exceedingly difficult. We propose FedFace, a federated learning (FL) framework for collaborative learning of face recognition models in a privacy-aware manner. FedFace utilizes the face images available on multiple clients to learn an accurate and generalizable face recognition model where the face images stored at each client are neither shared with other clients nor the central host and each client is a mobile device containing face images pertaining to only the owner of the device (one identity per client). Our experiments show the effectiveness of FedFace in enhancing the verification performance of pre-trained face recognition system on standard face verification benchmarks namely LFW, IJB-A, and IJB-C.

cs.CV↗

arXiv2021

Data-Free Knowledge Distillation for Heterogeneous Federated Learning

Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation} approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that, our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.

cs.LG↗

arXiv2021

MTNet: A Multi-Task Neural Network for On-Field Calibration of Low-Cost Air Monitoring Sensors

The advances of sensor technology enable people to monitor air quality through widely distributed low-cost sensors. However, measurements from these sensors usually encounter high biases and require a calibration step to reach an acceptable performance in down-streaming analytical tasks. Most existing calibration methods calibrate one type of sensor at a time, which we call single-task calibration. Despite the popularity of this single-task schema, it may neglect interactions among calibration tasks of different sensors, which encompass underlying information to promote calibration performance. In this paper, we propose a multi-task calibration network (MTNet) to calibrate multiple sensors (e.g., carbon monoxide and nitrogen oxide sensors) simultaneously, modeling the interactions among tasks. MTNet consists of a single shared module, and several task-specific modules. Specifically, in the shared module, we extend the multi-gate mixture-of-experts structure to harmonize the task conflicts and correlations among different tasks; in each task-specific module, we introduce a feature selection strategy to customize the input for the specific task. These improvements allow MTNet to learn interaction information shared across different tasks, and task-specific information for each calibration task as well. We evaluate MTNet on three real-world datasets and compare it with several established baselines. The experimental results demonstrate that MTNet achieves the state-of-the-art performance.

cs.LG↗

arXiv2021

Off-Policy Imitation Learning from Observations

Learning from Observations (LfO) is a practical reinforcement learning scenario from which many applications can benefit through the reuse of incomplete resources. Compared to conventional imitation learning (IL), LfO is more challenging because of the lack of expert action guidance. In both conventional IL and LfO, distribution matching is at the heart of their foundation. Traditional distribution matching approaches are sample-costly which depend on on-policy transitions for policy learning. Towards sample-efficiency, some off-policy solutions have been proposed, which, however, either lack comprehensive theoretical justifications or depend on the guidance of expert actions. In this work, we propose a sample-efficient LfO approach that enables off-policy optimization in a principled manner. To further accelerate the learning procedure, we regulate the policy update with an inverse action model, which assists distribution matching from the perspective of mode-covering. Extensive empirical results on challenging locomotion tasks indicate that our approach is comparable with state-of-the-art in terms of both sample-efficiency and asymptotic performance.

cs.LG↗

arXiv2021

Learning Deep Neural Networks under Agnostic Corrupted Supervision

Training deep neural models in the presence of corrupted supervision is challenging as the corrupted data points may significantly impact the generalization performance. To alleviate this problem, we present an efficient robust algorithm that achieves strong guarantees without any assumption on the type of corruption and provides a unified framework for both classification and regression problems. Unlike many existing approaches that quantify the quality of the data points (e.g., based on their individual loss values), and filter them accordingly, the proposed algorithm focuses on controlling the collective impact of data points on the average gradient. Even when a corrupted data point failed to be excluded by our algorithm, the data point will have a very limited impact on the overall loss, as compared with state-of-the-art filtering methods based on loss values. Extensive experiments on multiple benchmark datasets have demonstrated the robustness of our algorithm under different types of corruption.

cs.LG↗

arXiv2020

Robust Collaborative Learning with Noisy Labels

Learning with curriculum has shown great effectiveness in tasks where the data contains noisy (corrupted) labels, since the curriculum can be used to re-weight or filter out noisy samples via proper design. However, obtaining curriculum from a learner itself without additional supervision or feedback deteriorates the effectiveness due to sample selection bias. Therefore, methods that involve two or more networks have been recently proposed to mitigate such bias. Nevertheless, these studies utilize the collaboration between networks in a way that either emphasizes the disagreement or focuses on the agreement while ignores the other. In this paper, we study the underlying mechanism of how disagreement and agreement between networks can help reduce the noise in gradients and develop a novel framework called Robust Collaborative Learning (RCL) that leverages both disagreement and agreement among networks. We demonstrate the effectiveness of RCL on both synthetic benchmark image data and real-world large-scale bioinformatics data.

cs.LG↗

arXiv2020

Mimicking Full-Duplex Secure Communications for Buffer-Aided Multi-Relay Systems

This paper considers secure communication in buffer-aided cooperative wireless networks in the presence of one eavesdropper, which can intercept the data transmission from both the source and relay nodes. A new max-ratio relaying protocol is proposed, in which different relays are chosen for reception and transmission according to the ratio of the legitimate channels to the eavesdropper channels, so that the relay selected for reception and the relay selected for transmission can receive and transmit at the same time. It is worth noting that the relay employs a randomize-and-forward (RF) strategy such that the eavesdropper can only decode the signals received in the two hops independently. Theoretical analysis of the secrecy throughput of the proposed scheme is provided and the approximate closed-form expressions are derived, which are verified by simulations. Through numerical results, it is shown that the proposed scheme achieves a significant improvement in secrecy throughput compared with existing relay selection policies.

cs.IT↗

arXiv2020

Alternate Distributed Beamforming for Buffer-Aided Multi-Antenna Relay Systems

In this paper, link selection is investigated in half-duplex (HD) dual-hop cooperative systems with multiple antennas at the relays. Alternate distributed beamforming (ADB) scheme is revisited for buffer-aided multi-antenna relay systems, in which the relays are divided into two groups, with one group receiving the same information broadcast from the source and the other group transmitting the common messages to the destination via distributed beamforming in each time slot. It is worth noting that the relays used for reception and transmission are determined without the need of instantaneous channel state information (CSI). Theoretical analysis of the achievable throughput of the proposed scheme in Rayleigh fading is provided and the approximate closed-form expressions are derived. Simulation results are given to verify the theoretical analysis. Through numerical results, it is shown that compared with existing link selection policies, the fixed scheduling ADB scheme achieves a significant improvement in achievable throughput. It is also shown that for the ADB scheme, the throughput performance of increasing the number of antennas equipped at each relay is better than that of increasing the number of relays equipped with a single antenna when the total number of antennas at the relays is fixed.

cs.IT↗

arXiv2020

Learning Sparse Rewarded Tasks from Sub-Optimal Demonstrations

Model-free deep reinforcement learning (RL) has demonstrated its superiority on many complex sequential decision-making problems. However, heavy dependence on dense rewards and high sample-complexity impedes the wide adoption of these methods in real-world scenarios. On the other hand, imitation learning (IL) learns effectively in sparse-rewarded tasks by leveraging the existing expert demonstrations. In practice, collecting a sufficient amount of expert demonstrations can be prohibitively expensive, and the quality of demonstrations typically limits the performance of the learning policy. In this work, we propose Self-Adaptive Imitation Learning (SAIL) that can achieve (near) optimal performance given only a limited number of sub-optimal demonstrations for highly challenging sparse reward tasks. SAIL bridges the advantages of IL and RL to reduce the sample complexity substantially, by effectively exploiting sup-optimal demonstrations and efficiently exploring the environment to surpass the demonstrated performance. Extensive empirical results show that not only does SAIL significantly improve the sample-efficiency but also leads to much better final performance across different continuous control tasks, comparing to the state-of-the-art.

cs.LG↗

arXiv2019

Efficient Collaborative Multi-Agent Deep Reinforcement Learning for Large-Scale Fleet Management

Large-scale online ride-sharing platforms have substantially transformed our lives by reallocating transportation resources to alleviate traffic congestion and promote transportation efficiency. An efficient fleet management strategy not only can significantly improve the utilization of transportation resources but also increase the revenue and customer satisfaction. It is a challenging task to design an effective fleet management strategy that can adapt to an environment involving complex dynamics between demand and supply. Existing studies usually work on a simplified problem setting that can hardly capture the complicated stochastic demand-supply variations in high-dimensional space. In this paper we propose to tackle the large-scale fleet management problem using reinforcement learning, and propose a contextual multi-agent reinforcement learning framework including three concrete algorithms to achieve coordination among a large number of agents adaptive to different contexts. We show significant improvements of the proposed framework over state-of-the-art approaches through extensive empirical studies.

cs.MA↗

arXiv2019

Ranking Policy Gradient

Sample inefficiency is a long-lasting problem in reinforcement learning (RL). The state-of-the-art estimates the optimal action values while it usually involves an extensive search over the state-action space and unstable optimization. Towards the sample-efficient RL, we propose ranking policy gradient (RPG), a policy gradient method that learns the optimal rank of a set of discrete actions. To accelerate the learning of policy gradient methods, we establish the equivalence between maximizing the lower bound of return and imitating a near-optimal policy without accessing any oracles. These results lead to a general off-policy learning framework, which preserves the optimality, reduces variance, and improves the sample-efficiency. Furthermore, the sample complexity of RPG does not depend on the dimension of state space, which enables RPG for large-scale problems. We conduct extensive experiments showing that when consolidating with the off-policy learning framework, RPG substantially reduces the sample complexity, comparing to the state-of-the-art.

cs.LG↗

arXiv2019

Alternate Distributed Beamforming for Decode-and-Forward Multi-Relay Systems Using Buffers

In this paper, link selection for half-duplex buffer-aided relay systems is revisited. A new fixed scheduling policy referred as alternate distributed beamforming (ADB) is proposed, in which the relays are divided into two groups, with one group receiving the same information broadcast from the source and the other group transmitting the common messages to the destination via distributed beamforming in each time slot. It is worth noting that the relays used for reception and transmission are determined without the need of instantaneous channel state information (CSI). Theoretical analysis of the achievable throughput of the proposed scheme in Rayleigh fading is provided and the approximate closed-form expressions are derived. Through numerical results, it is shown that compared with existing link selection policies, the proposed fixed scheduling ADB achieves a significant improvement in achievable throughput.

cs.IT↗

arXiv2019

Model-Protected Multi-Task Learning

Multi-task learning (MTL) refers to the paradigm of learning multiple related tasks together. In contrast, in single-task learning (STL) each individual task is learned independently. MTL often leads to better trained models because they can leverage the commonalities among related tasks. However, because MTL algorithms can ``leak" information from different models across different tasks, MTL poses a potential security risk. Specifically, an adversary may participate in the MTL process through one task and thereby acquire the model information for another task. The previously proposed privacy-preserving MTL methods protect data instances rather than models, and some of them may underperform in comparison with STL methods. In this paper, we propose a privacy-preserving MTL framework to prevent information from each model leaking to other models based on a perturbation of the covariance matrix of the model matrix. We study two popular MTL approaches for instantiation, namely, learning the low-rank and group-sparse patterns of the model matrix. Our algorithms can be guaranteed not to underperform compared with STL methods. We build our methods based upon tools for differential privacy, and privacy guarantees, utility bounds are provided, and heterogeneous privacy budgets are considered. The experiments demonstrate that our algorithms outperform the baseline methods constructed by existing privacy-preserving MTL methods on the proposed model-protection problem.

stat.ML↗

arXiv2019

Inferring the Importance of Product Appearance: A Step Towards the Screenless Revolution

Nowadays, almost all the online orders were placed through screened devices such as mobile phones, tablets, and computers. With the rapid development of the Internet of Things (IoT) and smart appliances, more and more screenless smart devices, e.g., smart speaker and smart refrigerator, appear in our daily lives. They open up new means of interaction and may provide an excellent opportunity to reach new customers and increase sales. However, not all the items are suitable for screenless shopping, since some items' appearance play an important role in consumer decision making. Typical examples include clothes, dolls, bags, and shoes. In this paper, we aim to infer the significance of every item's appearance in consumer decision making and identify the group of items that are suitable for screenless shopping. Specifically, we formulate the problem as a classification task that predicts if an item's appearance has a significant impact on people's purchase behavior. To solve this problem, we extract features from three different views, namely items' intrinsic properties, items' images, and users' comments, and collect a set of necessary labels via crowdsourcing. We then propose an iterative semi-supervised learning framework with three carefully designed loss functions. We conduct extensive experiments on a real-world transaction dataset collected from the online retail giant JD.com. Experimental results verify the effectiveness of the proposed method.

cs.CV↗

arXiv2019

MetaPred: Meta-Learning for Clinical Risk Prediction with Limited Patient Electronic Health Records

In recent years, increasingly augmentation of health data, such as patient Electronic Health Records (EHR), are becoming readily available. This provides an unprecedented opportunity for knowledge discovery and data mining algorithms to dig insights from them, which can, later on, be helpful to the improvement of the quality of care delivery. Predictive modeling of clinical risk, including in-hospital mortality, hospital readmission, chronic disease onset, condition exacerbation, etc., from patient EHR, is one of the health data analytic problems that attract most of the interests. The reason is not only because the problem is important in clinical settings, but also there are challenges working with EHR such as sparsity, irregularity, temporality, etc. Different from applications in other domains such as computer vision and natural language processing, the labeled data samples in medicine (patients) are relatively limited, which creates lots of troubles for effective predictive model learning, especially for complicated models such as deep learning. In this paper, we propose MetaPred, a meta-learning for clinical risk prediction from longitudinal patient EHRs. In particular, in order to predict the target risk where there are limited data samples, we train a meta-learner from a set of related risk prediction tasks which learns how a good predictor is learned. The meta-learned can then be directly used in target risk prediction, and the limited available samples can be used for further fine-tuning the model performance. The effectiveness of MetaPred is tested on a real patient EHR repository from Oregon Health & Science University. We are able to demonstrate that with CNN and RNN as base predictors, MetaPred can achieve much better performance for predicting target risk with low resources comparing with the predictor trained on the limited samples available for this risk.

cs.LG↗