MSP-RNN: Multi-Step Piecewise Recurrent Neural Network for Predicting the Tendency of Services Invocation
MSP-RNN: Multi-Step Piecewise Recurrent Neural Network for Predicting the Tendency of Services Invocation
复制标题
MSP-RNN：用于预测服务调用趋势的多步分段循环神经网络
DOI：
10.1109/tsc.2020.2966487
复制
发表时间：
2022-03
影响因子：
8.1
通讯作者：
Bing Bai
中科院分区：
文献类型：
--
作者：
Haozhe Lin;YuShun Fan;Jia Zhang;Bing Bai
在面向服务的体系结构（SOA）的广泛应用的推动下，在过去的几十年中，越来越多的服务和混搭被开发并发布到Internet上。随着服务数量的不断增加，预测服务调用的趋势将为服务生态系统中的各种角色提供充满希望的机会。然而，服务调用具有三个独特的特点，这给预测带来了困难。首先，海量服务具有周期性、非线性和非平稳性等复杂的特性。第二，提供类似或补充功能的服务构成了错综复杂的关系。第三，mashup与其组成组件服务之间的组合依赖性进一步增加了难度。考虑到这些因素，我们开发了一个定制的模型多步分段递归神经网络（MSP-RNN）来预测服务调用的趋势。在MSP-RNN中，长短期记忆（LSTM）单元用于提取通用特征。基于这些特征，我们开发了一种分段回归机制来进行有区别的预测。此外，我们还开发了多步预测策略，以进一步提高预测精度和鲁棒性。在真实数据集上的大量实验表明，MSP-RNN可以更准确地预测服务调用的趋势，即，在对称平均绝对百分比误差（SMAPE）方面，比最先进的基线方法高出3.7%。
Driven by the widespread application of Service-Oriented Architecture (SOA), an increasing number of services and mashups have been developed and published onto the Internet in the past decades. With the number keeping on burgeoning, predicting the tendency of services invocation will provide various roles in service ecosystems with promising opportunities. However, services invocation bear three unique characteristics, which give rise to difficulties in predicting them. First, enormous services show different and complicated traits, like periodicity, nonlinearity and nonstationarity. Second, services providing similar or compensatory functions make up intricate relationship. Third, the combination dependencies between mashups and their comprising component services further amplify the difficulty. Given these factors, we have developed a tailored model Multi-Step Piecewise Recurrent Neural Network (MSP-RNN) to predict the tendency of services invocation. In MSP-RNN, Long Short Term Memory (LSTM) units are used to extract universal features. Based on these features, we have developed a piecewise regressive mechanism to make prediction discriminatingly. Besides, we have developed a multi-step prediction strategy to further enhance prediction accuracy and robustness. Extensive experiments in real-world data set with interpretable analysis show that MSP-RNN predicts the tendency of services invocation more accurately, i.e., by 3.7 percent in terms of symmetric mean absolute percentage error (SMAPE), than state-of-the-art baseline methods.
登录
查看更多内容
DOI：
10.1007/3-540-45065-3_8
发表时间：
2003-07
期刊：
Proceedings. IEEE Computer Society Bioinformatics Conference
影响因子：
--
作者：
M. Bicego;Vittorio Murino;Mário A. T. Figueiredo
通讯作者：
M. Bicego;Vittorio Murino;Mário A. T. Figueiredo
DOI：
10.1109/icws.2018.00013
发表时间：
2018-07
期刊：
2018 IEEE International Conference on Web Services (ICWS)
影响因子：
--
作者：
Haozhe Lin;Yushun Fan;Jia Zhang;Bing Bai
通讯作者：
Haozhe Lin;Yushun Fan;Jia Zhang;Bing Bai
影响因子：
64.8
作者：
RUMELHART, DE;HINTON, GE;WILLIAMS, RJ
通讯作者：
WILLIAMS, RJ
DOI：
10.1007/3-540-44864-0_30
发表时间：
2003-06
期刊：
--
影响因子：
--
作者：
T. Trafalis;Huseyin Ince;M. B. Richman
通讯作者：
T. Trafalis;Huseyin Ince;M. B. Richman
DOI：
10.1109/icws.2010.58
发表时间：
2010-07
期刊：
2010 IEEE International Conference on Web Services
影响因子：
--
作者：
Manish Godse;U. Bellur;R. Sonar
通讯作者：
Manish Godse;U. Bellur;R. Sonar