Search

Your search keyword '"Tang, Bin"' showing total 351 results
Start Over Author "Tang, Bin" Publisher ieee
351 results on '"Tang, Bin"'

27. Research of Deep Learning-Based Visibility Prediction Model for Foggy Days in Airport

Author

Wang Xichen, Tang Bin, Liu Cheng, and Wu Qian

Subjects
Polynomial regression, 010504 meteorology & atmospheric sciences, business.industry, Computer science, Deep learning, Process (computing), Regression analysis, 02 engineering and technology, Atmospheric model, 01 natural sciences, Wind speed, 020204 information systems, Statistics, Linear regression, 0202 electrical engineering, electronic engineering, information engineering, Artificial intelligence, business, Visibility, 0105 earth and related environmental sciences
Abstract

In this paper, we estimate the visibility distance under heavy fog conditions based on the airport video image containing information about the varying process of fog and the observed meteorological data. This article uses data from 00:00:16 to 11:47:48 on March 13, 2020, for an airport. By establishing two regression models, the Multiple Linear Regression (MLR) model and the Multiple Polynomial Regression (MPR) model, to analyzing the relationship between visibility distance and meteorological observation data on the ground, and comparing and evaluating the two models. Experimentally, it was found that MPR models solve relational equations with higher precision, and if it used a higher order of MPR model, it has higher predicted precision. And we establish a deep learning model based on video visibility distance estimation from video data and meteorological observation data of an airport and evaluate the accuracy of the estimated visibility. The experimental results show that the percentage error between the predicted value and the true value is less than 0.25%, which achieves high predictive accuracy and has a very good prediction effect.

Published
2021
Full Text
View/download PDF

28. Partial Synchronization to Accelerate Federated Learning Over Relay-Assisted Edge Networks.

Author

Qu, Zhihao, Guo, Song, Wang, Haozhao, Ye, Baoliu, Wang, Yi, Zomaya, Albert Y., and Tang, Bin

Subjects
LINEAR network coding, SYNCHRONIZATION, MACHINE learning
Abstract

Federated Learning (FL) is a promising machine learning paradigm to cooperatively train a global model with highly distributed data located on mobile devices. Aiming to optimize the communication efficiency for gradient aggregation and model synchronization among large-scale devices, we propose a relay-assisted FL framework. By breaking the traditional transmission-order constraint and exploiting the broadcast characteristic of relay nodes, we design a novel synchronization scheme named Partial Synchronization Parallel (PSP), in which models and gradients are transmitted simultaneously and aggregated at relay nodes, resulting in traffic reduction. We prove that PSP has the same convergence rate as the sequential synchronization approaches via rigorous analysis. To further accelerate the training process, we integrate PSP with any unbiased and error-bounded compression technologies and prove that the convergence properties of the resulting scheme still hold. Extensive experiments are conducted in a distributed cluster environment with real-world datasets and the results demonstrate that our proposed approach reduces the training time up to 37 percent compared to state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

37. Heterogeneity-Aware Gradient Coding for Tolerating and Leveraging Stragglers.

Author

Wang, Haozhao, Guo, Song, Tang, Bin, Li, Ruixuan, Yang, Yutong, Qu, Zhihao, and Wang, Yi

Subjects
ADAPTIVE computing systems, MACHINE learning, PARALLEL algorithms
Abstract

Distributed gradient descent has been widely adopted in the machine learning field because considerable computing resources are available when facing the huge volume of data. Specifically, the gradient over the whole data is cooperatively computed by multiple workers. However, its performance can be severely affected by slow workers, namely stragglers. Recently, coding-based approaches have been introduced to mitigate the straggler problem, but they could hardly deal with the heterogeneity among workers. Besides, they always discard the results of stragglers causing huge resource waste. In this article, we first investigate how to tolerate stragglers by discarding their results and then seek to leverage the stragglers. For tolerating stragglers, we propose a heterogeneity-aware coding scheme that encodes gradients adaptive to the computing capability of workers. Theoretically, this scheme is optimal for stragglers tolerance. Relying on the scheme, we further propose an algorithm called DHeter-aware to exploit the gradients of stragglers which we called delayed gradients. Moreover, theoretical results characterized for DHeter-aware exhibits the same convergence rate as the gradient descent without delayed gradients. Experiments on various tasks and clusters demonstrate that our coding scheme outperforms all the state-of-the-art methods and the DHeter-aware further accelerates the coding scheme by achieving 25 percent time savings. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results