OLD NEWs

  • Oct. 2020 | A new co-authored paper “Learning Tensor Train Representation With Automatic Rank Determination From Incomplete Noisy Data” has been submitted for possible publication (available online at arXiv). Congratulations to Le for his first paper submission.

  • Oct. 2020 | I will visit Prof. Ya-Feng Liu in Academy of Mathematics and Systems Science, Chinese Academy of Sciences from Oct. 9 to Oct. 25, 2020. Thanks very much for Ya-Feng's host.

  • Sep. 2020 | Cheers ! A co-authored journal paper “Adversarial Learning With Multi-Modal Attention for Visual Question Answering” was accepted by IEEE TNNLS (available online at IEEE Xplore ). Two co-authored journal papers were accepted by IEEE IOTJ and JCIN respectively.

  • Sep. 2020 | A new journal paper ‘‘Towards Probabilistic Tensor Canonical Polyadic Decomposition 2.0: Automatic Tensor Rank Learning Using Generalized Hyperbolic Prior" has been submitted for possible publication (available online at arXiv).

    • By noticing that the performance of tensor rank learning is related to the flexibility of the prior distribution, we introduced the generalized hyperbolic prior to the probabilistic modeling of the CPD problem, based on which an inference algorithm is further developed. Extensive numerical results showed that the proposed method exhibits remarkable performance in learning both low and high tensor ranks, even when the noise power is large.

    • This is my first time working with Prof. Sergios Theodoridis. Because of his books Pattern Recognition and Machine learning: A Bayesian and Optimization Perspective, I started to learn the knowledge of Bayesian learning (when I was a student). Thanks very much for his continuous education and guidance.

  • Sep. 2020 | Invited by Program Chair of IEEE HPCC 2020, I will serve on the technical program committee (TPC) for IEEE HPCC 2020.

  • Jul. 2020 | Invited by IEEE ICC 2021 WC Symposium Co-Chairs, I will serve on the technical program committee (TPC) for IEEE ICC 2021.

  • Jun. 2020 | I helped to organize a special session “Structured Matrix/Tensor Decompositions: Models, Applications and Fast Algorithms” in IEEE SAM 2020.