Filters








124 Hits in 1.1 sec

Federated f-Differential Privacy [article]

Qinqing Zheng, Shuxiao Chen, Qi Long, Weijie J. Su
2021 arXiv   pre-print
Federated learning (FL) is a training paradigm where the clients collaboratively learn models by repeatedly sharing information without compromising much on the privacy of their local sensitive data. In this paper, we introduce federated f-differential privacy, a new notion specifically tailored to the federated setting, based on the framework of Gaussian differential privacy. Federated f-differential privacy operates on record level: it provides the privacy guarantee on each individual record
more » ... f one client's data against adversaries. We then propose a generic private federated learning framework PriFedSync that accommodates a large family of state-of-the-art FL algorithms, which provably achieves federated f-differential privacy. Finally, we empirically demonstrate the trade-off between privacy guarantee and prediction performance for models trained by PriFedSync in computer vision tasks.
arXiv:2102.11158v1 fatcat:zbnhfqulpjck7bywdko4ntkvni

Learning Torque Control for Quadrupedal Locomotion [article]

Shuxiao Chen, Bike Zhang, Mark W. Mueller, Akshara Rai, Koushil Sreenath
2022 arXiv   pre-print
Reinforcement learning (RL) is a promising tool for developing controllers for quadrupedal locomotion. The design of most learning-based locomotion controllers adopts the joint position-based paradigm, wherein a low-frequency RL policy outputs target joint positions that are then tracked by a high-frequency proportional-derivative (PD) controller that outputs joint torques. However, the low frequency of such a policy hinders the advancement of highly dynamic locomotion behaviors. Moreover,
more » ... mining the PD gains for optimal tracking performance is laborious and dependent on the task at hand. In this paper, we introduce a learning torque control framework for quadrupedal locomotion, which trains an RL policy that directly predicts joint torques at a high frequency, thus circumventing the use of PD controllers. We validate the proposed framework with extensive experiments where the robot is able to both traverse various terrains and resist external pushes, given user-specified commands. To our knowledge, this is the first attempt of learning torque control for quadrupedal locomotion with an end-to-end single neural network that has led to successful real-world experiments among recent research on learning-based quadrupedal locomotion which is mostly position-based.
arXiv:2203.05194v1 fatcat:zqk7mmhhfvf2hcrueskhrse2je

Minimax Rates and Adaptivity in Combining Experimental and Observational Data [article]

Shuxiao Chen, Bo Zhang, Ting Ye
2021 arXiv   pre-print
A conceptually similar phase transition phenomenon has been identified in other statistical problems that involve multiple data sources [Chen et al., 2021] .  ... 
arXiv:2109.10522v1 fatcat:hmvvcns5encq3kyeaals7jfcwa

A Theorem of the Alternative for Personalized Federated Learning [article]

Shuxiao Chen, Qinqing Zheng, Qi Long, Weijie J. Su
2021 arXiv   pre-print
Our notion of heterogeneity is also related to the hierarchical Bayesian model considered in Bai et al. [2020] , Lucas et al. [2020] , Konobeev et al. [2020] , and Chen et al. [2020].  ...  Chen, S. Liu, and Z. Ma. Global and individualized community detection in inhomogeneous multilayer networks. arXiv preprint arXiv:2012.00933, 2020. G. Denevi, C. Ciliberto, D. Stamos, and M. Pontil.  ... 
arXiv:2103.01901v1 fatcat:4jvgbwklerbobhkgwplrgcpu2q

One-Way Matching of Datasets with Low Rank Signals [article]

Shuxiao Chen, Sizun Jiang, Zongming Ma, Garry P. Nolan, Bokai Zhu
2022 arXiv   pre-print
We study one-way matching of a pair of datasets with low rank signals. Under a stylized model, we first derive information-theoretic limits of matching. We then show that linear assignment with projected data achieves fast rates of convergence and sometimes even minimax rate optimality for this task. The theoretical error bounds are corroborated by simulated examples. Furthermore, we illustrate practical use of the matching procedure on two single-cell data examples.
arXiv:2204.13858v1 fatcat:yt6zmao23bchjlo4vrzubnmywu

Valid Inference Corrected for Outlier Removal [article]

Shuxiao Chen, Jacob Bien
2019 arXiv   pre-print
Ordinary least square (OLS) estimation of a linear regression model is well-known to be highly sensitive to outliers. It is common practice to (1) identify and remove outliers by looking at the data and (2) to fit OLS and form confidence intervals and p-values on the remaining data as if this were the original data collected. This standard "detect-and-forget" approach has been shown to be problematic, and in this paper we highlight the fact that it can lead to invalid inference and show how
more » ... ntly developed tools in selective inference can be used to properly account for outlier detection and removal. Our inferential procedures apply to a general class of outlier removal procedures that includes several of the most commonly used approaches. We conduct simulations to corroborate the theoretical results, and we apply our method to three real data sets to illustrate how our inferential results can differ from the traditional detect-and-forget strategy. A companion R package, outference, implements these new procedures with an interface that matches the functions commonly used for inference with lm in R.
arXiv:1711.10635v3 fatcat:7beydwpsrrc5dpo2xqzakyg2na

Weighted Training for Cross-Task Learning [article]

Shuxiao Chen, Koby Crammer, Hangfeng He, Dan Roth, Weijie J. Su
2022 arXiv   pre-print
In this paper, we introduce Target-Aware Weighted Training (TAWT), a weighted training algorithm for cross-task learning based on minimizing a representation-based task distance between the source and target tasks. We show that TAWT is easy to implement, is computationally efficient, requires little hyperparameter tuning, and enjoys non-asymptotic learning-theoretic guarantees. The effectiveness of TAWT is corroborated through extensive experiments with BERT on four sequence tagging tasks in
more » ... ural language processing (NLP), including part-of-speech (PoS) tagging, chunking, predicate detection, and named entity recognition (NER). As a byproduct, the proposed representation-based task distance allows one to reason in a theoretically principled way about several critical aspects of cross-task learning, such as the choice of the source data and the impact of fine-tuning.
arXiv:2105.14095v2 fatcat:h6trz3ezbfccvmmusmwvbvz6ju

Real-time Geo-localization Using Satellite Imagery and Topography for Unmanned Aerial Vehicles [article]

Shuxiao Chen, Xiangyu Wu, Mark W. Mueller, Koushil Sreenath
2021 arXiv   pre-print
The capabilities of autonomous flight with unmanned aerial vehicles (UAVs) have significantly increased in recent times. However, basic problems such as fast and robust geo-localization in GPS-denied environments still remain unsolved. Existing research has primarily concentrated on improving the accuracy of localization at the cost of long and varying computation time in various situations, which often necessitates the use of powerful ground station machines. In order to make image-based
more » ... calization online and pragmatic for lightweight embedded systems on UAVs, we propose a framework that is reliable in changing scenes, flexible about computing resource allocation and adaptable to common camera placements. The framework is comprised of two stages: offline database preparation and online inference. At the first stage, color images and depth maps are rendered as seen from potential vehicle poses quantized over the satellite and topography maps of anticipated flying areas. A database is then populated with the global and local descriptors of the rendered images. At the second stage, for each captured real-world query image, top global matches are retrieved from the database and the vehicle pose is further refined via local descriptor matching. We present field experiments of image-based localization on two different UAV platforms to validate our results.
arXiv:2108.03344v1 fatcat:onja4qa3fzaazaqipbgo2ydp4i

Chemerin-9 Attenuates Experimental Abdominal Aortic Aneurysm Formation in ApoE−/− Mice

Shuxiao Chen, Chenglin Han, Shuai Bian, Jianfeng Chen, Xuedong Feng, Gang Li, Xuejun Wu, Sumanta Chatterjee
2021 Journal of Oncology  
Authors' Contributions Shuxiao Chen and Gang Li conceived and designed the experiments; Shuxiao Chen, Gang Li, Chenglin Han, Jianfeng Chen, and Shuai Bian performed the experiments; Shuxiao Chen, Xuedong  ...  Feng, and Chenglin Han analyzed and interpreted the data; Xuejun Wu and Gang Li contributed reagents, materials, and analysis tools; Shuxiao Journal of Oncology 13 Chen and Chenglin Han wrote the manuscript  ... 
doi:10.1155/2021/6629204 pmid:33953746 pmcid:PMC8068550 fatcat:3ql5sr56bjc5lcxdtgl7wif4zu

Molecular and Histologic Adaptation of Horned Gall Induced by the Aphid Schlechtendalia chinensis (Pemphigidae)

Qin Lu, Xiaoming Chen, Zixiang Yang, Nawaz Haider Bashir, Juan Liu, Yongzhong Cui, Shuxiao Shao, Ming-Shun Chen, Hang Chen
2021 International Journal of Molecular Sciences  
Chinese galls are the result of hyperplasia in host plants induced by aphids. The metabolism and gene expression of these galls are modified to accommodate the aphids. Here, we highlight the molecular and histologic features of horned galls according to transcriptome and anatomical structures. In primary pathways, genes were found to be unevenly shifted and selectively expressed in the galls and leaves near the galls (LNG). Pathways for amino acid synthesis and degradation were also unevenly
more » ... fted, favoring enhanced accumulation of essential amino acids in galls for aphids. Although galls enhanced the biosynthesis of glucose, which is directly available to aphids, glucose content in the gall tissues was lower due to the feeding of aphids. Pathways of gall growth were up-regulated to provide enough space for aphids. In addition, the horned gall has specialized branched schizogenous ducts and expanded xylem in the stalk, which provide a broader feeding surface for aphids and improve the efficiency of transportation and nutrient exchange. Notably, the gene expression in the LNG showed a similar pattern to that of the galls, but on a smaller scale. We suppose the aphids manipulate galls to their advantage, and galls lessen competition by functioning as a medium between the aphids and their host plants.
doi:10.3390/ijms22105166 pmid:34068250 pmcid:PMC8153119 fatcat:pmaqtobgbjcrrfc5mzew66ayua

Learning from Multi-User Multi-Attribute Annotations [chapter]

Ou Wu, Shuxiao Li, Honghui Dong, Ying Chen, Weiming Hu
2014 Proceedings of the 2014 SIAM International Conference on Data Mining  
Chen et al. [4] leveraged a conditional random field (CRF) to model attribute correlations with the aim of improving the classification on each attribute.  ... 
doi:10.1137/1.9781611973440.3 dblp:conf/sdm/WuLDCH14 fatcat:ldjzpbqpcvavnogqi2o4ikp5e4

Label-Aware Neural Tangent Kernel: Toward Better Generalization and Local Elasticity [article]

Shuxiao Chen, Hangfeng He, Weijie J. Su
2020 arXiv   pre-print
., 2018; Ji & Telgarsky, 2019; Chen et al., 2019) . Limitations of the NTK and corrections.  ...  Zixiang Chen, Yuan Cao, Difan Zou, and Quanquan Gu. How much over-parameterization is sufficient to learn deep relu networks? arXiv preprint arXiv:1911.12360, 2019.  ... 
arXiv:2010.11775v2 fatcat:6fd7jfn465eqhf4zjdgwlsjooq

Vision-aided Dynamic Quadrupedal Locomotion on Discrete Terrain using Motion Libraries [article]

Ayush Agrawal, Shuxiao Chen, Akshara Rai, Koushil Sreenath
2022 arXiv   pre-print
In this paper, we present a framework rooted in control and planning that enables quadrupedal robots to traverse challenging terrains with discrete footholds using visual feedback. Navigating discrete terrain is challenging for quadrupeds because the motion of the robot can be aperiodic, highly dynamic, and blind for the hind legs of the robot. Additionally, the robot needs to reason over both the feasible footholds as well as robot velocity by speeding up and slowing down at different parts of
more » ... the terrain. We build an offline library of periodic gaits which span two trotting steps on the robot, and switch between different motion primitives to achieve aperiodic motions of different step lengths on an A1 robot. The motion library is used to provide targets to a geometric model predictive controller which controls stance. To incorporate visual feedback, we use terrain mapping tools to build a local height map of the terrain around the robot using RGB and depth cameras, and extract feasible foothold locations around both the front and hind legs of the robot. Our experiments show a Unitree A1 robot navigating multiple unknown, challenging and discrete terrains in the real world.
arXiv:2110.00891v2 fatcat:hfjvfr5jjfh3hhnm5pg7bieteq

Feedback Control for Autonomous Riding of Hovershoes by a Cassie Bipedal Robot [article]

Shuxiao Chen, Jonathan Rogers, Bike Zhang, Koushil Sreenath
2019 arXiv   pre-print
Motivated towards achieving multi-modal locomotion, in this paper, we develop a framework for a bipedal robot to dynamically ride a pair of Hovershoes over various terrain. Our developed control strategy enables the Cassie bipedal robot to interact with the Hovershoes to balance, regulate forward and rotational velocities, achieve fast turns, and move over flat terrain, slopes, stairs, and rough outdoor terrain. Our sensor suite comprising of tracking and depth cameras for visual SLAM as well
more » ... our Dijkstra-based global planner and timed elastic band-based local planning framework enables us to achieve autonomous riding on the Hovershoes while navigating an obstacle course. We present numerical and experimental validations of our work.
arXiv:1907.11353v2 fatcat:k6rdszwahvczzl7ttqai4qizuq

CD36 deficiency affects depressive-like behaviors possibly by modifying gut microbiota and the inflammasome pathway in mice

Shunjie Bai, Wei Wang, Ting Wang, Juan Li, Shuxiao Zhang, Zhi Chen, Xunzhong Qi, Jianjun Chen, Ke Cheng, Peng Xie
2021 Translational Psychiatry  
AbstractBoth inflammatory processes and gut microbiota have been implicated in the pathophysiology of depressive disorders. The class B scavenger receptor CD36 is involved in the cytotoxicity associated with inflammation. However, its role in depression has not yet been examined. In this study, we investigated whether CD36 affects depression by modulating the microbiota-gut-inflammasome-brain axis. We used CD36−/− (knockout) mice subjected to chronic social defeat stress, and measured the
more » ... sion of CD36 in these depressed mice and in patients with depression. The hippocampus of CD36−/− mice was used to investigate changes in the NLRP3 inflammasome signaling pathway. The 16S rRNA gene sequence-based approach was used to compare the cecal microbial communities in CD36−/− and WT mice. The CD36 deficiency in CD36−/− mice alleviated chronic stress-induced depression-like behaviors. CD36 was upregulated in depressed mice as well as in depressed patients. Furthermore, the NLRP3 inflammasome signaling pathway was downregulated in the hippocampus of CD36−/− mice. The Simpson Diversity Index revealed increased cecal bacterial alpha-diversity in the CD36−/− mice. Among genera, Bacteroides, Rikenella, and Alloprevotella were significantly more abundant in the CD36−/− mice, whereas Allobaculum was less abundant, consistent with the attenuated inflammation in the hippocampus of CD36−/− mice. Our findings suggest that CD36 deficiency changes the gut microbiota composition, which in turn may impact depressive-like behaviors by affecting the inflammasome pathway.
doi:10.1038/s41398-020-01130-8 pmid:33414380 fatcat:3o67lk3fmbfnjbb7cmer2ndj3i
« Previous Showing results 1 — 15 out of 124 results