報(bào)告人: 王建軍 教授
講座日期：2020-11-19
講座時(shí)間：15:00
報(bào)告地點(diǎn)：騰訊會(huì)議（768 415 831）
主辦單位：數(shù)學(xué)與信息科學(xué)學(xué)院
講座人簡(jiǎn)介：
王建軍，西南大學(xué)教授，巴渝學(xué)者特聘教授，重慶市創(chuàng)新創(chuàng)業(yè)領(lǐng)軍人才，重慶工業(yè)與應(yīng)用數(shù)學(xué)學(xué)會(huì)副理事長(zhǎng)，CSIAM全國(guó)大數(shù)據(jù)與人工智能專家委員會(huì)委員，美國(guó)數(shù)學(xué)評(píng)論評(píng)論員，曾獲重慶市自然科學(xué)獎(jiǎng)勵(lì)三等獎(jiǎng)。主要研究方向?yàn)椋焊呔S數(shù)據(jù)建模與挖掘、深度學(xué)習(xí)、壓縮感知與張量恢復(fù)、函數(shù)逼近論等。在神經(jīng)網(wǎng)絡(luò)(深度學(xué)習(xí))逼近復(fù)雜性和高維數(shù)據(jù)稀疏建模等方面有一定的學(xué)術(shù)積累。多次出席國(guó)際、國(guó)內(nèi)重要學(xué)術(shù)會(huì)議，并應(yīng)邀做大會(huì)特邀報(bào)告22余次。已在IEEE Transactions on Pattern Analysis and Machine Intelligence, Applied and Computational Harmonic Analysis, Inverse Problems, Neural Networks, Signal Processing, IEEE Signal Processing letters, Journal of Computational and Applied Mathematics，ICASSP，中國(guó)科學(xué)(A、F輯), 數(shù)學(xué)學(xué)報(bào), 計(jì)算機(jī)學(xué)報(bào)，電子學(xué)報(bào)等知名專業(yè)期刊發(fā)表90余篇學(xué)術(shù)論文。主持國(guó)家自然科學(xué)基金5項(xiàng)，教育部科學(xué)技術(shù)重點(diǎn)項(xiàng)目1項(xiàng)，重慶市自然科學(xué)基金1項(xiàng)，主研8項(xiàng)國(guó)家自然、社會(huì)科學(xué)基金；現(xiàn)主持國(guó)家自然科學(xué)基金面上項(xiàng)目2項(xiàng)，參與國(guó)家重點(diǎn)基礎(chǔ)研究發(fā)展973計(jì)劃1項(xiàng)。
講座簡(jiǎn)介：
This talk focuses on the recovery of low-tubal-rank tensors from binary measurements based on tensor-tensor product (or t-product) and tensor Singular Value Decomposition (t-SVD). Two types of recovery models are considered; one is the tensor hard singular tube thresholding and the other is the tensor nuclear-norm minimization. In the case no random dither exists in the measurements, our research shows that the direction of tensor $\mathcal{X} \in \R^{n_1\times n_2\times n_3}$ with tubal rank r can be well approximated from $\Omega((n_1+n_2)n_3r)$ random Gaussian measurements. In the case nonadaptive adaptive dither exists in the measurements, it is proved that both the direction and the magnitude of $\mathcal{X}$ can be simultaneously recovered. As we will see, under the nonadaptive adaptive measurement scheme, the recovery errors of two reconstruction procedures decay at the rate of polynomial of the oversampling factor $\lambda:=m/(n_1+n_2)n_3r$,i.e., $\mathcal{O}(\lambda^{-1/6})$ and $\mathcal{O}(\lambda^{-1/4})$, respectively. In order to obtain faster decay rate, we introduce a recursive strategy and allow the dithers in quantization adaptive to previous measurements for each iterations. Under this quantization scheme, two iterative recovery algorithms are proposed which establish recovery errors decaying at the rate of exponent of the oversampling factor, i.e., $\exp(-\mathcal{O}(\lambda))$. Numerical experiments on both synthetic and real-world data sets are conducted and demonstrate the validity of our theoretical results and the superiority of our algorithms.
