【摘要】：為提高服務(wù)運(yùn)行質(zhì)量,需要主動(dòng)預(yù)防服務(wù)失效和服務(wù)性能波動(dòng),而不是在服務(wù)發(fā)生錯(cuò)誤時(shí)觸發(fā)處理程序。高效地預(yù)測(cè)與分析大規(guī)模服務(wù)的性能是有效可行的主動(dòng)預(yù)防工具。然而傳統(tǒng)的服務(wù)性能預(yù)測(cè)模型多采用完全批量訓(xùn)練模式,難以滿足大規(guī)模服務(wù)計(jì)算的實(shí)時(shí)性要求。在綜合權(quán)衡完全批量學(xué)習(xí)法和隨機(jī)梯度下降法的基礎(chǔ)上,建立了基于在線學(xué)習(xí)的大規(guī)模服務(wù)性能預(yù)測(cè)模型,提出了一種基于小批量在線學(xué)習(xí)的服務(wù)性能預(yù)測(cè)方法,通過(guò)合理地設(shè)置預(yù)測(cè)模型的批量參數(shù),一次迭代僅需訓(xùn)練批量規(guī)模較小的樣本數(shù)據(jù),從而改善大規(guī)模服務(wù)性能預(yù)測(cè)的時(shí)間效率;詳細(xì)分析了在線服務(wù)預(yù)測(cè)模型的收斂性。實(shí)驗(yàn)表明,提出的在線學(xué)習(xí)算法有效地解決了大規(guī)模服務(wù)預(yù)測(cè)算法的時(shí)效性問(wèn)題。
[Abstract]:In order to improve the quality of service, it is necessary to actively prevent service failure and service performance fluctuations, rather than trigger a handler when a service error occurs. Efficient prediction and analysis of the performance of large-scale services is an effective and feasible active prevention tool. However, the traditional service performance prediction model mostly adopts the full batch training model, which is difficult to meet the real-time requirements of large-scale service computing. On the basis of synthetically weighing the complete batch learning method and the stochastic gradient descent method, a large-scale service performance prediction model based on online learning is established, and a service performance prediction method based on small batch online learning is proposed. By reasonably setting the batch parameters of the prediction model, only the sample data of small batch size need to be trained in one iteration to improve the time efficiency of large-scale service performance prediction, and the convergence of online service prediction model is analyzed in detail. Experiments show that the proposed online learning algorithm can effectively solve the time-efficiency problem of large-scale service prediction algorithm.
【作者單位】： 南京航空航天大學(xué)計(jì)算機(jī)科學(xué)與技術(shù)學(xué)院;滁州學(xué)院地理信息科學(xué)系;上海第二工業(yè)大學(xué)計(jì)算機(jī)與信息學(xué)院;
【基金】：國(guó)家自然科學(xué)基金Nos.61672022,61272036 安徽省高等學(xué)校自然科學(xué)基金No.KJ2017A414~~
【分類號(hào)】：TP181

【相似文獻(xiàn)】

相關(guān)期刊論文 前3條

1 初元璋,祁鵬,張婭;寶鋼IF鋼大生產(chǎn)產(chǎn)品性能預(yù)測(cè)[J];北京科技大學(xué)學(xué)報(bào);2001年01期

2 王丹民;李華德;郝春輝;;邯鋼數(shù)據(jù)傳輸系統(tǒng)的建立及基于該系統(tǒng)的帶鋼性能預(yù)測(cè)[J];冶金自動(dòng)化;2006年03期

3 劉振宇;周思杭;譚建榮;唐宏亮;;基于多準(zhǔn)則修正的產(chǎn)品性能多參數(shù)關(guān)聯(lián)分析與預(yù)測(cè)方法[J];機(jī)械工程學(xué)報(bào);2013年15期

相關(guān)會(huì)議論文 前1條

1 陳超;李忠輝;熊昆鵬;;神經(jīng)網(wǎng)絡(luò)模型在沙鋼熱軋性能預(yù)測(cè)中的應(yīng)用[A];全國(guó)冶金自動(dòng)化信息網(wǎng)2014年會(huì)論文集[C];2014年

，

本文編號(hào)：2285633