本文關(guān)鍵詞： 大宗商品 價(jià)格預(yù)測(cè) 組合模型 可視分析　出處：《東華大學(xué)學(xué)報(bào)(自然科學(xué)版)》2017年04期 　論文類型：期刊論文

【摘要】：大宗商品價(jià)格因受國際和國內(nèi)眾多因素的影響而具有較大的波動(dòng)性,對(duì)其進(jìn)行準(zhǔn)確預(yù)測(cè)具有較大的挑戰(zhàn).從對(duì)大宗商品價(jià)格影響因素的篩選出發(fā),提出了基于因素分析的組合預(yù)測(cè)方法.對(duì)一年期的甲醇價(jià)格的跟蹤預(yù)測(cè)表明,以廣義自回歸條件異方差(generalized auto-regressive conditional heteroskedasticity,GARCH)模型和自回歸移動(dòng)平均(auto-regressive and moving average,ARMA)模型相結(jié)合的組合預(yù)測(cè)模型對(duì)甲醇價(jià)格的中長期趨勢(shì)預(yù)測(cè)有較好的效果.為結(jié)合專家的經(jīng)驗(yàn)判斷,彌補(bǔ)已有方法對(duì)波動(dòng)拐點(diǎn)預(yù)測(cè)滯后的不足,并對(duì)各類組合模型的預(yù)測(cè)效果進(jìn)行動(dòng)態(tài)比較,構(gòu)建了一個(gè)融合專家經(jīng)驗(yàn)值的動(dòng)態(tài)可視分析系統(tǒng).
[Abstract]:Commodity prices have great volatility because of the influence of many international and domestic factors, so it is a great challenge to predict them accurately. A combined forecasting method based on factor analysis is proposed. The combination of generalized auto-regressive conditional heteroscedastic auto-regressive conditional heteroscedastic auto-regressive model and autoregressive autoregressive and moving average ARMA model has a good effect on the medium and long term trend prediction of methanol price. This paper makes up for the deficiency of the existing methods to predict the inflection point of fluctuation, and makes a dynamic comparison of the forecasting effect of all kinds of combined models, and constructs a dynamic visual analysis system which integrates the experience value of experts.
【作者單位】： 華東師范大學(xué)經(jīng)濟(jì)管理學(xué)部;華東師范大學(xué)計(jì)算機(jī)科學(xué)與軟件工程學(xué)院;
【基金】：上海市科技發(fā)展基金軟科學(xué)研究資助項(xiàng)目(17692104400) 國家自然科學(xué)基金資助項(xiàng)目(61672237)
【分類號(hào)】：F713.35;F767
，

本文編號(hào)：1551251