AI security and Privacy Forum issue 11 - stable learning: finding common ground between causal reasoning and machine learning

Introduction to the Forum

Forum time

2022 year 03 month 03 Japan （ Thursday ）10:00-12:00

Forum agenda

01 Subject report

host ：

Wu Baoyuan ( Shenzhen associate professor of the Chinese University of Hong Kong ）

The guest ：

Cui Peng ( Associate professor of Tsinghua University )

Report title ：

Stable Learning: Finding the Common Ground between Causal Inference and Machine Learning

Steady learning : Look for common ground between causal reasoning and machine learning

( Already in Nature The journal published https://www.nature.com/articles/s42256-022-00445-z）

02 Tencent is credible AI Achievement sharing

host ：

Bian Yatao ( tencent AI Lab Senior researcher )

The guest ：

Wubingzhe ( tencent AI Lab Senior researcher )

Wang Huanchao ( Tencent Research Institute researcher )

03 Round Table Forum

host ：

Cao Jianfeng ( Senior researcher of Tencent Research Institute )

The guest ：

Cui Peng ( Associate professor of Tsinghua University )

Zhangjiyu ( Executive dean of the future rule of Law Research Institute of Renmin University of China )

Yao Jianhua ( tencent AI Lab AI Chief medical scientist )

issue ：

1. trusted AI The importance and value of

2. Explainable AI The status quo of 、 Challenges and legal requirements     

3. Interpretability and AI The realization of fairness

Sponsor unit

- Shenzhen big data research institute

- Chinese society of image graphics

To undertake unit

- tencent AI lab

- Tencent research institute

Co Organizer

- Chinese University of Hong Kong （ Shenzhen ） School of data science

- IEEE Guangzhou Section Biometrics Council Chapter

Report form  

Bili Bili live ：http://live.bilibili.com/22947067

Live video of Tencent Research Institute

About the reporter

Cui Peng （ Professor of Tsinghua University ）

Cui Peng , Associate professor of Tsinghua University . He is in 2010 He received a doctor's degree from Tsinghua University . His research interests include causal reasoning and stable learning 、 Network representation learning and social dynamics modeling . He's learning by machine 、 Published in famous conferences and journals in the field of data mining and multimedia 100 Many papers . His recent research has yielded 5 Best Paper Award , And were selected as 2014 Years and 2016 Year of 《KDD A special issue 》. He is IEEE TKDE, IEEE TBD, ACM TIST, ACM TOMM, DMKD and KAIS Deputy editor's, etc ,ACM CIKM19 and MMM2020 Co chair of the project . He is ACM and CCF Outstanding members of ,IEEE Senior member of .

Peng Cui is an Associate Professor with tenure in Tsinghua University. He got his PhD degree from Tsinghua University in 2010. His research interests include causal inference and stable learning, network representation learning, and social dynamics modeling. He has published more than 100 papers in prestigious conferences and journals in machine learning, data mining and multimedia. His recent research won 5 best paper awards and were selected into the Best of KDD special issues in 2014 and 2016 respectively. He is the Associate Editor of IEEE TKDE, IEEE TBD, ACM TIST, ACM TOMM, DMKD and KAIS etc., and the program co-chair of ACM CIKM19 and MMM2020. He is a Distinguished Member of ACM and CCF, and Senior Member of IEEE.

Report content

Stable Learning: Finding the Common Ground between Causal Inference and Machine Learning

Steady learning : Look for common ground between causal reasoning and machine learning

In a common machine learning problem , Models estimated using training data sets , Predict future outcome values based on observed characteristics . When test data and training data come from the same distribution , Many learning algorithms have been proposed and proved to be successful . However , For a given training data distribution , The best performing models often take advantage of subtle statistical relationships between features , This makes them useful when applied to data whose test distribution is different from the training data , May be more prone to prediction errors . How to develop a stable and robust learning model for data transfer , It is very important for academic research and practical application . Causal reasoning refers to the process of drawing a conclusion according to the conditions under which the causal relationship occurs , It is a powerful statistical modeling tool for interpretive and stable learning . In this speech , We will focus on the latest developments in stable learning , It aims to explore causal knowledge from observed data , To improve the interpretability and stability of machine learning algorithm .

Predicting future outcome values based on their observed features using a model estimated on a training data set in a common machine learning problem. Many learning algorithms have been proposed and shown to be successful when the test data and training data come from the same distribution. However, the best-performing models for a given distribution of training data typically exploit subtle statistical relationships among features, making them potentially more prone to prediction error when applied to test data whose distribution differs from that in training data. How to develop learning models that are stable and robust to shifts in data is of paramount importance for both academic research and real applications. Causal inference, which refers to the process of drawing a conclusion about a causal connection based on the conditions of the occurrence of an effect, is a powerful statistical modeling tool for explanatory and stable learning. In this talk, we focus on the latest progress of stable learning, aiming to explore causal knowledge from observational data to improve the interpretability and stability of machine learning algorithms. 

invite to one 's side men of wisdom and valor  

• Chinese University of Hong Kong ( Shenzhen )、 Shenzhen big data research institute recruits full-time research scientists in the direction of artificial intelligence security and privacy 、 Data Engineer 、 Interview students , And postdoctoral 、2022 Doctoral students enrolled in autumn ( Artificial intelligence security and privacy 、 Computer vision 、 Machine learning, etc ). More information about the position , Please click on Recruitment Links For more information .

• tencent AI Lab trusted AI The Technology Center recruits regular employees and research interns ( Fairness , Interpretability , Robust optimization ). More information about the position , Please scan the QR code below for more information .