Current Visitors:



Review of Seminars in 2023


The 9th BIGSCITY Seminar (2023.10.13)


The ninth BIGSCITY seminar was successfully held online and offline on October 13, 2023. The main content of this seminar is the introduction and discussion of papers. A total of three students will interpret and share the selected papers, and communicate with other participants on the content of the papers. The selected papers at this meeting are from KDD'23 and Arxiv, covering multiple directions such as regional representation learning, multivariate time series prediction, model fairness, and uncertainty research. read more


Jiawei Cheng:    Using OpenStreetMap building footprint for urban area representation learning


Jiahao Ji:             Research on fairness issues in multivariate time series prediction


Zehua Liu:           The Application of Uncertainty in Traffic Prediction

Video Playback:

The 8th BIGSCITY Seminar (2023.9.7)


The theme of this month's seminar is the interpretation of this year's KDD spatiotemporal related papers. Four students will choose their own research direction to interpret and discuss the relevant papers in the KDD conference. The content of the report covers transferable graph structure learning, spatiotemporal knowledge graph joint relationship evolution learning, localized adaptive spatiotemporal graph neural network, VAE based human trajectory synthesis, etc. read more


Dayan Pan:         TransGTR


Zimeng Li:          Co-evolutionary learning of spatiotemporal knowledge graph for small and medium-sized enterprise supply chain forecasting


Chengkai Han:  Localized adaptive spatiotemporal graph neural network


Yujing Lin:          Human trajectory synthesis based on VAE

The 7th BIGSCITY Seminar (2023.7.28)


The theme of this month's seminar is the interpretation of spatiotemporal papers related to KDD in recent years. Four students will choose relevant papers from the KDD conference for interpretation and discussion based on their research direction. The content of the report covers spatiotemporal prediction based on multi-step dependency, multivariate time series modeling and prediction, dynamic graph feature learning, graph neural networks and spatiotemporal prediction, etc. read more


Lida Guo:              Spatiotemporal prediction based on multi-step dependency relationships


Qiushi Feng:        Time series interpolation method based on location-aware graph and variational encoder


Wentao Zhang:   Spatiotemporal dynamic graph representation learning


Zhibo Zhang:       STEP: Enhanced multivariate time series prediction with pre-trained models and spatiotemporal graph neural networks

The 6th BIGSCITY Seminar (2023.6.16)


On June 16th, BIGSCITY Lab successfully held a special session on spatiotemporal data mining at the KDD Workshop. Five students who were accepted for KDD 2023 presented their research topics and engaged in discussions with the attendees. The presentations covered various topics including time series data prediction, scholar influence profiling, node popularity prediction, spatiotemporal causal learning, and spatiotemporal point processes. read more


Chen Yang:               Time series heterogeneity analysis with wavelet/dynamic time warping and attention mechanism


Yuankai Luo:            Scholar influence profiling based on self-citation graph


Shuo Ji:                      Node popularity prediction based on community structure and dynamic graph representation learning


Zhengyang Zhou:   Invariant correlation learning for spatiotemporal data distribution outliers

The 5th BIGSCITY Seminar (2023.5.17)


In this spatiotemporal data mining paper seminar, five students admitted to KDD2023 will discuss their research directions with everyone. The content of the report covers time series data prediction, scholar influence profiling, node popularity prediction, spatiotemporal causal learning, spatiotemporal point processes, etc. read more


Yichuan Zhang:   Traditional machine learning methods and VC theory in high-dimensional data


Xuanhao Shi:       Methods and comparisons for distributed training models


Xiaohan Jiang:    Self-supervised learning for time series


Chen Yang:          Deep learning in finance

The 4th BIGSCITY Seminar (2023.4.5)


On April 5th, the BIGSCITY Lab conducted its fourth spatiotemporal data mining workshop, where four students engaged in discussions regarding their respective research topics. The presentations covered spatiotemporal data mining, transfer learning, causal inference, representation learning, and trajectory quality enhancement.  read more


Yujing Lin:             Interdisciplinary research on operations research and transportation science


Wentao Zhang:   Causal and spatiotemporal data mining


Yifan Yang:           Urban area representation learning


Yudong Li:            Survey on trajectory quality enhancement techniques

The 3rd BIGSCITY Seminar (2023.3.5)


On March 5th, BIGSCITY Lab hosted its third spatiotemporal data mining paper workshop, with four students discussing their research topics with the audience. The presentations encompassed spatiotemporal data mining, AI in healthcare, infectious disease modeling, traffic forecasting, and federated learning. read more


Chengkai Han:     Application of federated learning in spatiotemporal domain


Xuanhao Shi:        Spatiotemporal data and AI in infectious disease research


Dayan Pan:           Data patterns in electronic health records


Yu Mou:                 Enhanced traffic prediction using graph decomposition

The 2nd BIGSCITY Seminar (2023.1)



Jiawei Cheng:      Attack on spatiotemporal prediction models based on saliency values


Qiushi Feng:        Time series prediction modeling and self-supervised learning


Xiaohan Jiang:    Time-frequency analysis and applications in time series


Peiyu Wang:         Epidemic prediction model integrating multiple data sources