應(yīng)yl7703永利官網(wǎng)李周平教授和趙學(xué)靖副教授邀請，廣州大學(xué)經(jīng)濟與統(tǒng)計學(xué)院胡建明副教授將于2021年12月23日訪問我校，期間將舉辦專題學(xué)術(shù)報告。

報告題目：Study on the monotone quantile prediction model and its application

報告時間：12月23日（星期四）下午4：00

報告地點：理工樓631報告廳

摘要：Multiple quantile simultaneous prediction is conducive to providing more uncertainty information to decision-makers for risk quantification.In this talk, a monotone quantile regression neural network (MQRNN) framework is presentedfor time series quantile forecasting. The proposed approach takes the monotonicity of quantile into consideration and handle the quantile crossing problem by adding the quantile information into the input structure and using the gradient based point-wise loss function. In view of the complex characteristics of time series, such as time-varying and asymmetric heavy-tail features, a new quantile function is utilized to describe the complete conditional distribution information of data. Under this model framework, non-crossing multiple quantiles can be estimated and predicted simultaneously. The proposed approach is implemented based on the common neural network architecture, and the constructed model are applied to actual data in different fields. The results show that the constructed models can provide accurate and reliable multi-quantile prediction results, and alleviate the problem of quantile crossing.

歡迎廣大師生光臨!

報告人簡介

胡建明，理學(xué)博士，廣州大學(xué)經(jīng)濟與統(tǒng)計學(xué)院副教授，入選廣州大學(xué)“百人計劃”，碩士生導(dǎo)師，加拿大皇后大學(xué)訪問學(xué)者。主要研究興趣為應(yīng)用統(tǒng)計，機器學(xué)習(xí)及統(tǒng)計方法在能源領(lǐng)域的應(yīng)用。在國際期刊《Renewable & Sustainable Energy Reviews》，《Energy Conversion and Management》和《Applied Soft Computing》等雜志發(fā)表SCI論文近20篇（SCI一區(qū)17篇），發(fā)表論文的總被為引次數(shù)為978次，單篇最高引用次數(shù)為165次，兩篇入選ESI高被引論文。主持或完成國家自然科學(xué)基金項目2項，廣東省自然科學(xué)基金面上項目1項，廣州市基金項目1項。

甘肅應(yīng)用數(shù)學(xué)中心

甘肅省高校應(yīng)用數(shù)學(xué)與復(fù)雜系統(tǒng)省級重點實驗室

蘭州大學(xué)大數(shù)據(jù)科學(xué)研究中心

yl7703永利官網(wǎng)

萃英學(xué)院

二〇二一年十二月二十二日