Parallel Computing of Spatio-Temporal Model Based on Deep Reinforcement Learning

被引:4
|
作者
Lv, Zhiqiang [1 ,2 ]
Li, Jianbo [1 ,2 ]
Xu, Zhihao [1 ]
Wang, Yue [1 ]
Li, Haoran [1 ]
机构
[1] Qingdao Univ, Coll Comp Sci Technol, Qingdao 266071, Peoples R China
[2] Inst Ubiquitous Networks & Urban Comp, Qingdao 266070, Peoples R China
基金
中国国家自然科学基金;
关键词
Parallel computing methodologies; Deep learning; Reinforcement learning; Gradient accumulation algorithm;
D O I
10.1007/978-3-030-85928-2_31
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Deep learning parallel plays an important role in accelerating model training and improving prediction accuracy. In order to fully consider the authenticity of the simulation application scenario of model, the development of deep learning model is becoming more complex and deeper. However, a more complex and deeper model requires a larger amount of computation compared to common spatio-temporal model. In order to speed up the calculation speed and accuracy of the deep learning model, this work optimizes the common spatial-temporal model in deep learning from three aspects: data parallel, model parallel and gradient accumulation algorithm. Firstly, the data parallel slicing algorithm proposed in this work achieves parallel GPUs load balancing. Secondly, this work independently parallelizes the components of the deep spatio-temporal. Finally, this work proposes a gradient accumulation algorithm based on deep reinforcement learning. This work uses two data sets (GeoLife and Chengdu Taxi) to train and evaluate multiple parallel modes. The parallel mode combining data parallel and gradient accumulation algorithm is determined. The experimental effect has been greatly improved compared with the original model.
引用
收藏
页码:391 / 403
页数:13
相关论文
共 50 条
  • [41] Spatio-temporal fragmentation-aware time-varying service provisioning in computing power networks based on model-assisted reinforcement learning
    Ma, Huangxu
    Zhang, Jiawei
    Gu, Zhiqun
    Kilper, Daniel C.
    Ji, Yuefeng
    JOURNAL OF OPTICAL COMMUNICATIONS AND NETWORKING, 2023, 15 (11) : 788 - 803
  • [42] Action Recognition Based on Efficient Deep Feature Learning in the Spatio-Temporal Domain
    Husain, Farzad
    Dellen, Babette
    Torras, Carme
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2016, 1 (02): : 984 - 991
  • [43] Spatio-temporal prediction for distributed PV generation system based on deep learning neural network model
    Dai, Qiangsheng
    Huo, Xuesong
    Hao, Yuchen
    Yu, Ruiji
    FRONTIERS IN ENERGY RESEARCH, 2023, 11
  • [44] Modeling spatio-temporal distribution of soil moisture by deep learning-based cellular automata model
    Xiaodong Song
    Ganlin Zhang
    Feng Liu
    Decheng Li
    Yuguo Zhao
    Jinling Yang
    Journal of Arid Land, 2016, 8 : 734 - 748
  • [45] Modeling spatio-temporal distribution of soil moisture by deep learning-based cellular automata model
    SONG Xiaodong
    ZHANG Ganlin
    LIU Feng
    LI Decheng
    ZHAO Yuguo
    YANG Jinling
    Journal of Arid Land, 2016, 8 (05) : 734 - 748
  • [46] Modeling spatio-temporal distribution of soil moisture by deep learning-based cellular automata model
    Song Xiaodong
    Zhang Ganlin
    Liu Feng
    Li Decheng
    Zhao Yuguo
    Yang Jinling
    JOURNAL OF ARID LAND, 2016, 8 (05) : 734 - 748
  • [47] Sensor-Based Human Activity Recognition with Spatio-Temporal Deep Learning
    Nafea, Ohoud
    Abdul, Wadood
    Muhammad, Ghulam
    Alsulaiman, Mansour
    SENSORS, 2021, 21 (06) : 1 - 20
  • [48] Deep Latent Factor Model for Spatio-Temporal Forecasting
    Koo, Wonmo
    Ma, Eun-Yeol
    Kim, Heeyoung
    TECHNOMETRICS, 2024, 66 (03) : 470 - 482
  • [49] Multi-agent Deep Reinforcement Learning with Spatio-Temporal Feature Fusion for Traffic Signal Control
    Du, Xin
    Wang, Jiahai
    Chen, Siyuan
    Liu, Zhiyue
    MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021: APPLIED DATA SCIENCE TRACK, PT IV, 2021, 12978 : 470 - 485
  • [50] Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal Modeling
    Sosic, Adrian
    Zoubir, Abdelhak M.
    Rueckert, Elmar
    Peters, Jan
    Koeppl, Heinz
    JOURNAL OF MACHINE LEARNING RESEARCH, 2018, 19