Ziqi Liu
Greetings! I have been working at Ant Group since 2017. I was a visiting researcher advised by Prof. Alex Smola at Machine Learning Department, Carnegie Mellon University. I held a Ph.D degree in Department of Computer Science, Xi'an Jiaotong Univerity, under Prof. Qinghua Zheng.
Research Interests
My principal research interests lie in machine learning, especially focus on probabilistic models, graphical models, nonparametric modeling, large-scale inference algorithms and applications in user modeling, text mining:
Developing scalable inference for various kinds of statistical models from both algorithm and system perspectives. This means modifying the models, then boosting the algorithms on a single multi-cores/GPUs machine by fully exploiting system architectures, or distributing them on many machines with decent convergence.
Finding principled techniques to automatically solve the problems in terms of user modeling, text mining, temporal models by developing various structured probabilistic models, Bayesian Nonparametric models.
Research Experience
Ant Group. 2017~now
Carnegie Mellon University. (Visting researcher) Advisor: Alex Smola, 2014~2016
Xi'an Jiaotong University. (Research Assistant) Advisor: Qinghua Zheng, 2010~2017
Publications
MixSeq: Connecting Macroscopic Time Series Forecasting with Microscopic Time Series Data.
Zhibo Zhu, Ziqi Liu, Ge Jin, Zhiqiang Zhang, Lei Chen, Jun Zhou, Jianyong Zhou.
NeurIPS. 2021. PDF
Learning Representations of Inactive Users: A Cross Domain Approach with Graph Neural Networks.
Ziqi Liu, Yue Shen, Xiaocheng Cheng et al.
CIKM. 2021.
Bandit Sampler for Training Graph Neural Networks.
Ziqi Liu, Zhengwei Wu, Zhiqiang Zhang, Jun Zhou, Shuang Yang, Le Song, Yuan Qi.
NeurIPS. 2020. PDF
Model-Protected Multi-Task Learning.
Jian Liang, Ziqi Liu, Jiayu Zhou, Xiaoqian Jiang, Changshui Zhang, Fei Wang.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE. 2020.
Hubble: an Industrial System for Audience Expansion in Mobile Marketing.
Chenyi Zhuang, Ziqi Liu, Zhiqiang Zhang, Yeze Tan, Zhengwei Wu, Zhining Liu, Jianping Wei, Jinjie Gu, Guannan Zhang, Jun Zhou, Yuan Qi.
KDD, 2020.
AGL: a Scalable System for Industrial-purpose Graph Machine Learning.
Dalong Zhang, Xin Huang, Ziqi Liu, Zhiyang Hu, Xianzheng Song, Zhibang Ge, Zhiqiang Zhang, Lin Wang, Jun Zhou, Yang Shuang, Yuan Qi.
VLDB, 2020.
Graph Representation Learning for Merchant Incentive Optimization in Mobile Payment Marketing.
Ziqi Liu, Dong Wang, Qianyu Yu, Zhiqiang Zhang et al.
CIKM, 2019.
Uncovering Insurance Fraud Conspiracy with Network Learning.
Chen Liang, Ziqi Liu, Bin Liu, Jun Zhou, Xiaolong Li, Shuang Yang, Yuan Qi.
SIGIR (Short paper), 2019.
GeniePath: Graph Neural Networks with Adaptive Receptive Paths.
Ziqi Liu, Chaochao Chen, Longfei Li, Jun Zhou, Xiaolong Li, Le Song, Yuan Qi.
AAAI, 2019. PDF
Heterogeneous Graph Neural Networks for Malicious Account Detection.
Ziqi Liu, Chaochao Chen, Xinxing Yang, Jun Zhou, Xiaolong Li, Le Song.
CIKM, 2018.
Distributed Collaborative Hashing and its Applications in Ant Financial.
Chaochao Chen, Ziqi Liu, Peilin Zhao, Jun Zhou, Xiaolong Li.
KDD, 2018. PDF
Privacy Preserving Point-of-Interest Recommendation Using Decentralized Matrix Factorization.
Chaochao Chen, Ziqi Liu, Peilin Zhao, Jun Zhou, Xiaolong Li.
AAAI, 2018.
Neural Network-based Graph Embedding for Malicious Accounts Detection.
Ziqi Liu, Chaochao Chen, Jun Zhou, Xiaolong Li, Feng Xu, Tao Chen, Le Song.
CCS Poster, 2017.
Attributing Hacks.
Ziqi Liu, Alex Smola, Kyle Soska, Yu-Xiang Wang, Qinghua Zheng.
In International Conference on Artificial Intelligence and Statistics (AISTATS), 2017
PDF (Attributing Hacks with Survival Trend Filtering, Electronic Journal of Statistics).
Joint Hacking and Latent Hazard Rate Estimation.
Ziqi Liu, Alex Smola, Kyle Soska, Yu-Xiang Wang, Qinghua Zheng.
In NIPS 2016 Workshop on Interpret ML. PDF
Nonparametric models for Characterizing the Topical Communities in Social Network.
Ziqi Liu, Qinghua Zheng, Fei Wang, Buyue Qian.
Neurocomputing. 2016.
DiFacto - Distributed Factorization Machines.
Mu Li, Ziqi Liu, Alex Smola, Yu-Xiang Wang.
WSDM, 2016. (Best Paper Runner-up). PDF
Fast Differentially Private Matrix Factorization.
Ziqi Liu, Yu-Xiang Wang, Alex Smola.
Modeling Users' Adoption Behaviors with Social Selection and Influence.
Ziqi Liu, Fei Wang, Qinghua Zheng.
SDM, 2015. (Best SDM'15 Finalist). PDF
A Dynamic Nonparametric Model for Characterizing the Topical Communities in Social Streams.
Ziqi Liu, Qinghua Zheng, Fei Wang, Zhenhua Tian, Bo Li.
A Random Walk Approach to Selectional Preferences Based on Preference Ranking and Propagation.
Zhenhua Tian, Hengheng Xiang, Ziqi Liu, Qinghua Zheng.
Codes
FastCofi (github.com/dmlc/mf), one of the fastest collaborative filtering system on a single machine, featured with:
Cache-efficiency (pushing the data flow to cache level)
A fast SGLD algorithm with dense updates and bookkeeping
Differential privacy
An efficient pipeline design that can handle massive data as long as the disk can hold
nDTCM, nonparametric Dynamic Topical Community Model, which is developed to analyze/organize/represent the data in social streams. Since the topical communities may evolve over time, our model can automatically adapt the number of latent communities or topics, and capturing the dynamic semantic drifts.
Nonparametric hazard regression via trend filtering (github.com/ziqilau), featured with:
Dynamic parameter spaces
Nonparametric hazard regression in trend filtering framework.