報(bào) 告 人：孔新兵 教授

報(bào)告題目：Matrix Quantile Factor Model

報(bào)告時(shí)間：2023年7月18日（周二）上午9：30 

報(bào)告地點(diǎn)：靜遠(yuǎn)樓1709學(xué)術(shù)報(bào)告廳

主辦單位：數(shù)學(xué)研究院、數(shù)學(xué)與統(tǒng)計(jì)學(xué)院、科學(xué)技術(shù)研究院

報(bào)告人簡介：

      孔新兵，南京審計(jì)大學(xué)統(tǒng)計(jì)與數(shù)據(jù)科學(xué)學(xué)院教授，主要研究興趣為高頻、高維數(shù)據(jù)統(tǒng)計(jì)推斷與機(jī)器學(xué)習(xí)。主持國家自然科學(xué)基金3項(xiàng)，參與重點(diǎn)項(xiàng)目1項(xiàng)。在統(tǒng)計(jì)學(xué)頂級(jí)期刊和計(jì)量經(jīng)濟(jì)學(xué)頂級(jí)期刊發(fā)表論文22篇。獲第一屆統(tǒng)計(jì)科學(xué)技術(shù)進(jìn)步獎(jiǎng)等獎(jiǎng)項(xiàng)。擔(dān)任RMTA和《應(yīng)用概率統(tǒng)計(jì)》編委。

報(bào)告摘要：

      In this talk, I will introduce a matrix quantile factor model for matrix-valued data with a low-rank structure. We estimate the row and column factor spaces via minimizing the empirical check loss function over all panels. We show the estimates converge at rate $1/\min\{\sqrt{p_1p_2}, \sqrt{p_2T},$ $\sqrt{p_1T}\}$ in average Frobenius norm, where $p_1$, $p_2$ and $T$  are the row dimensionality, column dimensionality and length of the matrix sequence. This rate is faster than that of the quantile estimates via ``flattening the matrix model into a large vector model. Smoothed estimates are given and their central limit theorems are derived under some mild condition. We provide three consistent criteria to determine the pair of row and column factor numbers. Extensive simulation studies and an empirical study justify our theory.