The Experts below are selected from a list of 229800 Experts worldwide ranked by ideXlab platform
Rui Duan - One of the best experts on this subject based on the ideXlab platform.
-
learning from local to global an efficient Distributed Algorithm for modeling time to event data
Journal of the American Medical Informatics Association, 2020Co-Authors: Rui Duan, Mary Regina Boland, Howard H Chang, Hua Xu, Martijn J Schuemie, Jiayi Tong, Jason Liang, Jiang Bian, John H HolmesAbstract:OBJECTIVE We developed and evaluated a privacy-preserving One-shot Distributed Algorithm to fit a multicenter Cox proportional hazards model (ODAC) without sharing patient-level information across sites. MATERIALS AND METHODS Using patient-level data from a single site combined with only aggregated information from other sites, we constructed a surrogate likelihood function, approximating the Cox partial likelihood function obtained using patient-level data from all sites. By maximizing the surrogate likelihood function, each site obtained a local estimate of the model parameter, and the ODAC estimator was constructed as a weighted average of all the local estimates. We evaluated the performance of ODAC with (1) a simulation study and (2) a real-world use case study using 4 datasets from the Observational Health Data Sciences and Informatics network. RESULTS On the one hand, our simulation study showed that ODAC provided estimates nearly the same as the estimator obtained by analyzing, in a single dataset, the combined patient-level data from all sites (ie, the pooled estimator). The relative bias was <0.1% across all scenarios. The accuracy of ODAC remained high across different sample sizes and event rates. On the other hand, the meta-analysis estimator, which was obtained by the inverse variance weighted average of the site-specific estimates, had substantial bias when the event rate is <5%, with the relative bias reaching 20% when the event rate is 1%. In the Observational Health Data Sciences and Informatics network application, the ODAC estimates have a relative bias <5% for 15 out of 16 log hazard ratios, whereas the meta-analysis estimates had substantially higher bias than ODAC. CONCLUSIONS ODAC is a privacy-preserving and noniterative method for implementing time-to-event analyses across multiple sites. It provides estimates on par with the pooled estimator and substantially outperforms the meta-analysis estimator when the event is uncommon, making it extremely suitable for studying rare events and diseases in a Distributed manner.
-
learning from electronic health records across multiple sites a communication efficient and privacy preserving Distributed Algorithm
Journal of the American Medical Informatics Association, 2020Co-Authors: Rui Duan, Mary Regina Boland, Howard H Chang, Hua Xu, Christopher H Schmid, Christopher B Forrest, John H Holmes, Martijn J Schuemie, Jesse A Berlin, Jason H MooreAbstract:OBJECTIVES: We propose a one-shot, privacy-preserving Distributed Algorithm to perform logistic regression (ODAL) across multiple clinical sites. MATERIALS AND METHODS: ODAL effectively utilizes the information from the local site (where the patient-level data are accessible) and incorporates the first-order (ODAL1) and second-order (ODAL2) gradients of the likelihood function from other sites to construct an estimator without requiring iterative communication across sites or transferring patient-level data. We evaluated ODAL via extensive simulation studies and an application to a dataset from the University of Pennsylvania Health System. The estimation accuracy was evaluated by comparing it with the estimator based on the combined individual participant data or pooled data (ie, gold standard). RESULTS: Our simulation studies revealed that the relative estimation bias of ODAL1 compared with the pooled estimates was <3%, and the ratio of standard errors was <1.25 for all scenarios. ODAL2 achieved higher accuracy (with relative bias <0.1% and ratio of standard errors <1.05). In real data analysis, we investigated the associations of 100 medications with fetal loss during pregnancy. We found that ODAL1 provided estimates with relative bias <10% for 85% of medications, and ODAL2 has relative bias <10% for 99% of medications. For communication cost, ODAL1 requires transferring p numbers from each site to the local site and ODAL2 requires transferring (p×p+p) numbers from each site to the local site, where p is the number of parameters in the regression model. CONCLUSIONS: This study demonstrates that ODAL is privacy-preserving and communication-efficient with small bias and high statistical efficiency.
Xiaohong Nian - One of the best experts on this subject based on the ideXlab platform.
-
Distributed Algorithm Design for Nonsmooth Resource Allocation Problems
IEEE Transactions on Cybernetics, 2020Co-Authors: Zhenhua Deng, Xiaohong NianAbstract:This paper investigates resource allocation problems, where the cost functions of agents are nonsmooth and the decisions of agents are constrained by heterogeneous local constraints and network resource constraints. We design a Distributed subgradient-based Algorithm to achieve the optimal resource allocation. Moreover, we analyze the convergence of the Algorithm to the optimal solution. The Algorithm can solve resource allocation problems with strongly convex cost functions and weight-balanced digraphs, as well as resource allocation problems with strictly convex cost functions and connected undirected graphs. With the Algorithm, the decisions of all agents asymptotically converge to the optimal allocation. Simulation examples verify the effectiveness of the Algorithm.
John H Holmes - One of the best experts on this subject based on the ideXlab platform.
-
learning from local to global an efficient Distributed Algorithm for modeling time to event data
Journal of the American Medical Informatics Association, 2020Co-Authors: Rui Duan, Mary Regina Boland, Howard H Chang, Hua Xu, Martijn J Schuemie, Jiayi Tong, Jason Liang, Jiang Bian, John H HolmesAbstract:OBJECTIVE We developed and evaluated a privacy-preserving One-shot Distributed Algorithm to fit a multicenter Cox proportional hazards model (ODAC) without sharing patient-level information across sites. MATERIALS AND METHODS Using patient-level data from a single site combined with only aggregated information from other sites, we constructed a surrogate likelihood function, approximating the Cox partial likelihood function obtained using patient-level data from all sites. By maximizing the surrogate likelihood function, each site obtained a local estimate of the model parameter, and the ODAC estimator was constructed as a weighted average of all the local estimates. We evaluated the performance of ODAC with (1) a simulation study and (2) a real-world use case study using 4 datasets from the Observational Health Data Sciences and Informatics network. RESULTS On the one hand, our simulation study showed that ODAC provided estimates nearly the same as the estimator obtained by analyzing, in a single dataset, the combined patient-level data from all sites (ie, the pooled estimator). The relative bias was <0.1% across all scenarios. The accuracy of ODAC remained high across different sample sizes and event rates. On the other hand, the meta-analysis estimator, which was obtained by the inverse variance weighted average of the site-specific estimates, had substantial bias when the event rate is <5%, with the relative bias reaching 20% when the event rate is 1%. In the Observational Health Data Sciences and Informatics network application, the ODAC estimates have a relative bias <5% for 15 out of 16 log hazard ratios, whereas the meta-analysis estimates had substantially higher bias than ODAC. CONCLUSIONS ODAC is a privacy-preserving and noniterative method for implementing time-to-event analyses across multiple sites. It provides estimates on par with the pooled estimator and substantially outperforms the meta-analysis estimator when the event is uncommon, making it extremely suitable for studying rare events and diseases in a Distributed manner.
-
learning from electronic health records across multiple sites a communication efficient and privacy preserving Distributed Algorithm
Journal of the American Medical Informatics Association, 2020Co-Authors: Rui Duan, Mary Regina Boland, Howard H Chang, Hua Xu, Christopher H Schmid, Christopher B Forrest, John H Holmes, Martijn J Schuemie, Jesse A Berlin, Jason H MooreAbstract:OBJECTIVES: We propose a one-shot, privacy-preserving Distributed Algorithm to perform logistic regression (ODAL) across multiple clinical sites. MATERIALS AND METHODS: ODAL effectively utilizes the information from the local site (where the patient-level data are accessible) and incorporates the first-order (ODAL1) and second-order (ODAL2) gradients of the likelihood function from other sites to construct an estimator without requiring iterative communication across sites or transferring patient-level data. We evaluated ODAL via extensive simulation studies and an application to a dataset from the University of Pennsylvania Health System. The estimation accuracy was evaluated by comparing it with the estimator based on the combined individual participant data or pooled data (ie, gold standard). RESULTS: Our simulation studies revealed that the relative estimation bias of ODAL1 compared with the pooled estimates was <3%, and the ratio of standard errors was <1.25 for all scenarios. ODAL2 achieved higher accuracy (with relative bias <0.1% and ratio of standard errors <1.05). In real data analysis, we investigated the associations of 100 medications with fetal loss during pregnancy. We found that ODAL1 provided estimates with relative bias <10% for 85% of medications, and ODAL2 has relative bias <10% for 99% of medications. For communication cost, ODAL1 requires transferring p numbers from each site to the local site and ODAL2 requires transferring (p×p+p) numbers from each site to the local site, where p is the number of parameters in the regression model. CONCLUSIONS: This study demonstrates that ODAL is privacy-preserving and communication-efficient with small bias and high statistical efficiency.
Didier El Baz - One of the best experts on this subject based on the ideXlab platform.
-
A Distributed Algorithm for a Reconfigurable Modular Surface
2014Co-Authors: Didier El Baz, Benoît Piranda, Julien BourgeoisAbstract:A Distributed Algorithm is proposed in order to control block motion of a reconfigurable micro-electro-mechanical modular surface. The modular surface is designed to convey fragile and tiny micro-parts. The Distributed Algorithm solves a discrete trajectory optimization problem. In particular, the Algorithm computes the shortest path between two points of the modular surface using a strategy based on minimum hop count. The proposed method based on Distributed asynchronous iterative elections is scalable.
-
Distributed Resolution of a Trajectory Optimization Problem on a MEMS-based Reconfigurable Modular Surface
2013Co-Authors: Serge Tembo, Didier El BazAbstract:In this paper we propose a Distributed Algorithm to solve a discrete trajectory optimization problem that occurs in a micro-electro-mechanical based modular surface context. The method computes the shortest path between two points of the modular surface using a strategy based on minimum hop count. Our scalable approach is based on Distributed asynchronous iterative elections. A multithreaded Java Smart Blocks Simulator used to validate our Distributed Algorithm is presented and some results obtained with the simulator are commented on.
Wei Liu - One of the best experts on this subject based on the ideXlab platform.
-
ICT - Multi-resource allocation for LTE networks: Joint-optimality and Distributed Algorithm
ICT 2013, 2013Co-Authors: Chungang Yang, Min Sheng, Qin Liu, Wei LiuAbstract:Current schemes of radio resource allocation for LTE Networks can be summarized as either the centralized or the Distributed. In this paper, a hybrid multi-resource allocation scheme is presented from a network overall performance perspective, where evolved NodeBs (eNBs) are motivated to assist users to implement the joint bandwidth scheduling and power water-filling on the physical resource block (PRB). Closed-form bandwidth and power solution are derived with the joint-optimality property proved. Then, a Distributed iteration Algorithm of joint bandwidth scheduling and power water-filling on the physical resource block (PRB) is developed. Numerical results verify the proposed Distributed Algorithm, which can achieve a better guarantee of throughput and fairness.
-
a selection function based Distributed Algorithm for delay constraint least cost unicast routing
International Conference on Communications, 2003Co-Authors: Wei Liu, Wenjing Lou, Yuguang FangAbstract:It is well-known that Distributed delay-constrained least-cost (DCLC) unicast routing problem is NP-complete. In this paper we propose an efficient Distributed Algorithm, namely, selection function based DCLC (SF-DCLC), based on a novel selection function for the DCLC problem. The proposed SF-DCLC Algorithm requires limited network state information at each network node and is always able to find a loop-free path satisfying the delay bound if such paths exist. Simulation study show that the SF-DCLC is not as sensitive to the delay bound and network size as some other DCLC routing Algorithms, and attains very low cost-inefficiency (less than 3% to the optimal One) in various network scenarios we simulate. The most attractive feature of SF-DCLC is that SF-DCLC has very high probability to find the optimal solution or a near-optimal solution in polynomial time with low computational complexity and message complexity.
-
ICC - A selection function based Distributed Algorithm for delay-constraint least-cost unicast routing
IEEE International Conference on Communications 2003. ICC '03., 2003Co-Authors: Wei Liu, Wenjing Lou, Yuguang FangAbstract:It is well-known that Distributed delay-constrained least-cost (DCLC) unicast routing problem is NP-complete. In this paper we propose an efficient Distributed Algorithm, namely, selection function based DCLC (SF-DCLC), based on a novel selection function for the DCLC problem. The proposed SF-DCLC Algorithm requires limited network state information at each network node and is always able to find a loop-free path satisfying the delay bound if such paths exist. Simulation study show that the SF-DCLC is not as sensitive to the delay bound and network size as some other DCLC routing Algorithms, and attains very low cost-inefficiency (less than 3% to the optimal One) in various network scenarios we simulate. The most attractive feature of SF-DCLC is that SF-DCLC has very high probability to find the optimal solution or a near-optimal solution in polynomial time with low computational complexity and message complexity.