Ziyu Zhou

Ziyu Zhou 周子渔

I am an MPhil student of the CityMind Lab, @HKUST(GZ), under the supervision of Prof. Yuxuan Liang. Previously, I completed my undergraduate degree in Computer Science and Technology at the Beijing University of Technology, where I conducted my research under the guidance of Prof. Gengyu Lyu at the DMS Lab.

During my undergraduate studies, I had the privilege of collaborating with Yiming Huang @HKUST(GZ) and Zihao Wang @HKUST. I aspire to build deep and enduring collaborations with fellow researchers in the future.

My research interests are structured around two principal domains:

Self-Supervised Learning for Time Series:
- Generative-based Methods: Focusing on diffusion-based generation.
- Adversarial Attacks and Robust Analysis: Examining the security and reliability of models.
- Advanced Pretraining Techniques: Utilizing data scaling laws and data distillation to enhance model training.
Multi-Modality Deep Learning:
- Explainability of Transformer-based models: Improving understanding of how models interpret and process text and image information.
- Application of Vision-Language Models on Streaming Data: Exploring the use of these models on dynamic inputs such as video.

These areas collectively aim to deepen our understanding of self-supervised learning of time series and enhance interpretability in multi-modal deep learning, pushing the boundaries of what these technologies can achieve.

Email1 / Email2 / Scholar / X / Github / CV

My Quote

"The purpose of computing is insight, not numbers." —— Richard Wesley Hamming

News

2024/09/02	A new life at HKUST(GZ) has begun! Feel free to say hi if we cross paths on campus!
2024/06/29	I was awarded as Outstanding Graduate Thesis and Outstanding Graduates of Beijing.
2024/04/17	One paper about time series forecasting was accepted by IJCAI'24.
2023/12/01	I decided to accept the MPhil offer from HKUST(GZ) Fok Ying Tung Graduate School.
2023/10/26	One paper about PM2.5 forecasting was accepted by Atmosphere.
2023/08/25	One paper about ASD detection was accepted by PRICAI'23.
2023/03/24	I achieved a satisfactory academic IELTS band score of 7.5 (8,8,6.5,6.5).

Research Programs & Publication

(* denotes corresponding author and ^ indicates equal contribution.)

	SDformer: Transformer with Spectral Filter and Dynamic Attention for Multivariate Time Series Long-term Forecasting Ziyu Zhou, Gengyu Lyu, Yiming Huang, Zihao Wang, Ziyu Jia, Zhen Yang The 33rd International Joint Conference on Artificial Intelligence (IJCAI '24, CCF-A & CORE-A) The Only Long Oral Paper of the Time Series Session (1/12) Paper Link / PDF / Code / Poster We propose a novel Transformer architecture (named SDformer) for long-term time series forecasting. It is the first time to address the problem of smooth attention distribution when modeling time series data with a large number of variates.
	TimesNet-PM2.5: Interpretable TimesNet for Disentangling Intraperiod and Interperiod Variations in PM2.5 Prediction Yiming Huang^, Ziyu Zhou^, Zihao Wang^, Xiaoying Zhi, Xiliang Liu* Atmosphere (JCR-Q3) Paper Link In this paper, we accomplish task-specific adaption of TimesNet (ICLR '23) named TimesNet-PM2.5. This specialized version improved the performance and interpretability of the PM2.5 prediction of Haikou, Hainan Province.
	STFM: Enhancing Autism Spectrum Disorder Classification Through Ensemble Learning-Based Fusion of Temporal and Spatial fMRI Patterns Ziyu Zhou^, Yiming Huang^, Yining Wang^, Yin Liang* The 20th Pacific Rim International Conference on Artificial Intelligence (PRICAI '23, CCF-C & CORE-B) Paper Link / PDF We propose a Spatial and Temporal framework named STFM based on cross-attention for better autism spectrum disorder classification. The results show that STFM's classification accuracy surpassed that of many machine learning models.
	CoC-GAN: Employing Context Cluster for Unveiling a New Pathway in Image Generation Zihao Wang^, Yiming Huang^, Ziyu Zhou^ Arxiv(2023) 2308.11857 arXiv We employ Context-Clustering Block (ICLR '23) into GAN for better interpretability.