forked from causal-machine-learning/kdd2023-workshop
-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
187 lines (142 loc) · 16.6 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
---
layout: home
search_exclude: true
image: images/logo.png
---
# **Causal Inference and Machine Learning in Practice**: Use cases for Product, Brand, Policy and Beyond
## **Schedule**
* Long Beach Convention & Entertainment Center, 300 E Ocean Blvd, Long Beach, CA 90802
([Map](https://goo.gl/maps/1N3XGEovGgJqXAV98))
* Date: August 7, 2023 (Monday)
* Time: 1:00 - 5:00 PM Pacific Time
## **Abstract**
The increasing demand for data-driven decision-making has led to the rapid growth of machine learning applications in
various industries. However, the ability to draw causal inferences from observational data remains a crucial challenge.
In recent years, causal inference has emerged as a powerful tool for understanding the effects of interventions in
complex systems. Combining causal inference with machine learning has the potential to provide a deeper understanding of
the underlying mechanisms and to develop more effective solutions to real-world problems.
This workshop aims to bring together researchers and practitioners from academia and industry to share their experiences
and insights on applying causal inference and machine learning techniques to real-world problems in the areas of
product, brand, policy, and beyond. The workshop welcomes original research that covers machine learning theory, deep
learning, causal inference, and online learning. Additionally, the workshop encourages topics that address scalable
system design, algorithm bias, and interpretability.
Through keynote talks, panel discussions, and contributed talks and posters, the workshop will provide a forum for
discussing the latest advances and challenges in applying causal inference and machine learning to real-world problems.
The workshop will also offer opportunities for networking and collaboration among researchers and practitioners working
in industry, government, and academia.
## **Paper Submission**
Please submit your paper to the [CMT portal](https://cmt3.research.microsoft.com/CMLKDD2023) site, and check the [Call for
Paper](https://causal-machine-learning.github.io/kdd2023-workshop/cfp/) page for details on important dates and
submission guidelines.
## **Outline**
| **Title** | **Speaker** | **Time (Duration)** | Link |
|-----------|-------------|--------------|------|
| **Introduction** | Organizers | 1:00 - 1:10 PM (10 minutes) | |
| **Invited Talk:** COG: Creative Optimality Gap for Video Advertising | [Raif Rustamov](#raif-rustamov-amazon) (Amazon) | 1:10 - 1:30 PM (20 minutes) | [Slides](https://drive.google.com/file/d/1ehHzNj-EDlhpQzCOhlJ9oE7Lnmf2Fpmc/view?usp=sharing)|
| **Invited Talk:** The Value of Last-Mile Delivery in Online Retail | [Ruomeng Cui](#ruomeng-cui-emory-university) (Emory) | 1:30 - 1:50 PM (20 minutes) | [Slides](https://drive.google.com/file/d/19w3ay80K1xBqceZj6DgBtyJgPx2OrYe6/view?usp=drive_link)|
| Leveraging Causal Uplift Modeling for Budget Constrained Benefits Allocation | Dmitri Goldenberg, Javier Albert (Booking.com) | 1:50 - 2:05 PM (15 minutes) | [Slides](https://docs.google.com/presentation/d/1Fz720lBj8DDsviLYNAIyBLVE7TN8qFoVtxPzbfAvo_I/edit?usp=drive_link)|
| Ensemble Method for Estimating Individualized Treatment Effects | Kevin Wu Han, Han Wu (Stanford) | 2:05 - 2:20 PM (15 minutes) | [Slides](https://drive.google.com/file/d/1KLAFpKYw5mlJQ7cNyLamS1LkuGScABSb/view?usp=sharing), [Paper](https://drive.google.com/file/d/1dzsIGgQ4cF2ltetH16A4Zz-L-pZ6tHBK/view?usp=drive_link) |
| A Scalable and Debiased Approach to Dynamic Pricing with Causal Machine Learning and Optimization | Nicolò Cosimo Albanese, Fabian Furrer, Marco Guerriero (Amazon AWS) | 2:20 - 2:35 PM (15 minutes) | [Slides](https://drive.google.com/file/d/1uSqpV51qxNcEyCrgzhYo2nr44tfn717z/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1VlW-zgrCfaKi5CtGYkQbkhWw7JVXhPtw/view?usp=drive_link) |
| An IPW-based Unbiased Ranking Metric in Two-sided Markets | Keisho Oh, Naoki Nishimura (Recruit Co), Minje Sung, Ken Kobayashi, Kazuhide Nakata (Tokyo Institute of Technology) | 2:35 - 2:50 PM (15 minutes) | [Slides](https://drive.google.com/file/d/1XLQCMUNy79jmYb0y1AB7ImRflq4csh6Y/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1Bshr6dwFB-E2H2K64g9OfERNUQcfP0eb/view?usp=drive_link)|
| **Break & Poster Session** | | 3:00 - 3:30 PM (30 minutes) | |
| **Invited Talk:** Unit Selection Based on Counterfactual Logic | [Ang Li](#ang-li-university-of-california-los-angeles) (UCLA) | 3:30 - 3:50 PM (20 minutes) | [Slides](https://drive.google.com/file/d/1P-it4MNrYbnWNUgodagU69oubVtpo_WV/view?usp=drive_link)|
| **Invited Talk:** Towards Automating the Causal Machine Learning Pipeline | [Vasilis Syrgkanis](#vasilis-syrgkanis-stanford-universityeconml) (Stanford/EconML) | 3:50 - 4:10 PM (20 minutes) | [Slides](https://www.dropbox.com/scl/fi/w2p1cnghhqp1qc377o4yu/auto_debiased.pptx?rlkey=8k4bdcrastqmzlao8n5bng258&dl=0)|
| Power and Pre-treatment Fit: Optimizing Synthetic Control Method for Quasi-experiments | Ali O Polat (Shipt) | 4:10 - 4:25 PM (15 minutes) | [Slides](https://drive.google.com/file/d/1F9bCmoCYg7AeNxKXekCze7a4ELXs2Ohv/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1rDsCmwl23HELiD-P_Qq9TUZZ_3opbEji/view?usp=drive_link)|
| Dynamic Causal Structure Discovery and Causal Effect Estimation | Jianian Wang, Rui Song (NCSU) | 4:25 - 4:40 PM (15 minutes) | [Slides](https://drive.google.com/file/d/13MzjyZLGxfGNoAg3TMRF0Es-TqtnRVSH/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1928QcX3PdFan_gkeJl-mxEgl7kQ9KupN/view?usp=drive_link)|
| Hierarchical Clustering As a Novel Solution to the Notorious Multicollinearity Problem in Observational Causal Inference | Yufei Wu, Zhiying Gu, Alex Deng, Jacob Zhu (Airbnb) | 4:40 - 4:55 PM (15 minutes) | [Slides](https://drive.google.com/file/d/1r0xXFFflSVDFNtGEypUGh845CeycF51q/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1HQekrFF1vNNrO7TF4aX43850-Dc6UHgo/view?usp=drive_link)|
## **Invited Speakers**
### Raif Rustamov, Amazon
#### Title: COG: Creative Optimality Gap for Video Advertising
#### Bio
Raif Rustamov is a Senior Applied Scientist at Amazon where he focuses on brand advertising science including relevance
modeling, representation learning, and causal inference. He previously worked as a Principal Inventive Scientist in AI
and Data Science at AT&T Labs conducting research on recommender systems, customer segmentation, identity for
cross-device advertising, and location analytics. Raif has a PhD in Applied and Computational Mathematics from Princeton
University and has taught at Purdue and Drew Universities, as well as worked as a research associate at Stanford
University.
#### Abstract
Video creatives play a crucial role in shaping consumer experiences and brand perceptions, but quantifying their impact
on shopper experience remains a complex challenge. In this talk, we introduce the Creative Optimality Gap (COG),
a metric developed to assess the relative optimality of video creatives using causal-inferential machine learning
methodology. Our main contributions include the development of the COG metric through the use of conditional individual
treatment effects projected on interpretable video features, the introduction of a meta-learner for its computation,
and the incorporation of model uncertainty to avoid false positives. Our work advances the understanding of video
creative effectiveness and provides a valuable tool for optimizing ad performance.
### Ruomeng Cui, Emory University
#### Title: The Value of Last-Mile Delivery in Online Retail
#### Bio
Ruomeng Cui is an Associate Professor of Operations Management at the Goizueta Business School, Emory University (on leave). She currently is a full-time Amazon Visiting Academic at Amazon, working in the supply chain domain. Her research focuses on causal inference, machine learning and data-driven modeling, with applications in retail, supply chains, and platforms. She currently serves as an associate editor for Manufacturing & Service Operations Management and Production and Operations Management. She received her Ph.D. in Operations Management from the Kellogg School of Management, Northwestern University and B.Sc in Industrial Engineering from Tsinghua University.
#### Abstract
Last-mile delivery has become increasingly important in the online retail industry. In this study, we study the economic
value of last-mile delivery. To do so, we conducted a quasi-experiment in collaboration with Cainiao, Alibaba's
logistics subsidiary, where home delivery was launched at some pickup stations in 2021. This allowed us to
comprehensively evaluate the causal impact of last-mile delivery. Using a difference-in-differences identification
method, we found that last-mile delivery significantly increases sales and customer spending on the retail platform. To
optimally prioritize limited delivery capacity, we employed causal machine learning to target the most responsive
customers. Our findings suggest that online retailers should carefully weigh the costs and benefits of last-mile
delivery and tailor their logistic strategies accordingly.
### Ang Li, University of California, Los Angeles
#### Title: Unit Selection Based on Counterfactual Logic
#### Bio
Dr. Li is set to join the Florida State University Department of Computer Science as an assistant professor in August.
He is currently a post-doctoral researcher in the Department of Computer Science at UCLA under the guidance of Prof.
Judea Pearl. His primary research area is causal inference, artificial intelligence, and causality-based
decision-making, with a focus on building causal models that estimate treatment effects (interventions) and evaluating
what would have happened if an individual had taken a treatment (counterfactuals). He is also interested in
decision-making modeling using knowledge of treatment effects and counterfactuals. Prior to his post-doc, Dr. Li
obtained his Ph.D. at UCLA with Prof. Judea Pearl and his M.S. degree at the University of Minnesota Twin Cities.
#### Abstract
The unit selection problem aims to identify a set of individuals who are most likely to
exhibit a desired mode of behavior, which is defined in counterfactual terms. A typical
example is that of selecting individuals who would respond one way if encouraged and a
different way if not encouraged. Unlike previous works on this problem, which rely on ad-hoc
heuristics, we approach this problem formally, using counterfactual logic, to properly capture
the nature of the desired behavior. This formalism enables us to derive an informative
selection criterion which integrates experimental and observational data. We show that a
more accurate selection criterion can be achieved when structural information is available
in the form of a causal diagram. We further discuss data availability issue regarding the
derivation of the selection criterion without the observational or experimental data. We
demonstrate the superiority of this criterion over A/B-test-based approaches.
### Vasilis Syrgkanis, Stanford University/EconML
#### Title: Towards Automating the Causal Machine Learning Pipeline
#### Bio
Vasilis Syrgkanis is an Assistant Professor in Management Science and Engineering and (by courtesy) in Computer Science,
in the School of Engineering at Stanford University. His research interests are in the areas of machine learning, causal
inference, econometrics, online and reinforcement learning, game theory/mechanism design and algorithm design. Until
August 2022, he was a Principal Researcher at Microsoft Research, New England, where he was a member of the EconCS and
StatsML groups. During his time at Microsoft, he co-led the project on Automated Learning and Intelligence for Causation
and Economics (ALICE) and was a co-founder of EconML, an open-source python package for causal machine learning. He
received his Ph.D. in Computer Science from Cornell University.
#### Abstract
## **Accepted Papers**
### For Oral Presentation
1. Leveraging Causal Uplift Modeling for Budget Constrained Benefits Allocation, Dmitri Goldenberg (Booking.com)*; Javier Albert (Booking.com); [Slides](https://docs.google.com/presentation/d/1Fz720lBj8DDsviLYNAIyBLVE7TN8qFoVtxPzbfAvo_I/edit?usp=drive_link)
2. Ensemble Method for Estimating Individualized Treatment Effects, Kevin Wu Han (Stanford University)*; Han Wu (Stanford University); [Slides](https://drive.google.com/file/d/1KLAFpKYw5mlJQ7cNyLamS1LkuGScABSb/view?usp=sharing), [Paper](https://drive.google.com/file/d/1dzsIGgQ4cF2ltetH16A4Zz-L-pZ6tHBK/view?usp=drive_link)
3. A Scalable and Debiased Approach to Dynamic Pricing with Causal Machine Learning and Optimization, Nicolò Cosimo Albanese (AWS)*; Fabian Furrer (AWS); Marco Guerriero (AWS); [Slides](https://drive.google.com/file/d/1uSqpV51qxNcEyCrgzhYo2nr44tfn717z/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1VlW-zgrCfaKi5CtGYkQbkhWw7JVXhPtw/view?usp=drive_link)
4. An IPW-based Unbiased Ranking Metric in Two-sided Markets, Keisho Oh (Recruit Co., Ltd.)*; Naoki Nishimura (Recruit Co., Ltd.); Minje Sung (Tokyo Institute of Technology); Ken Kobayashi (Tokyo Institute of Technology); Kazuhide Nakata (Department of Industrial Engineering and Economics, Tokyo Institute of Technology.); [Slides](https://drive.google.com/file/d/1XLQCMUNy79jmYb0y1AB7ImRflq4csh6Y/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1Bshr6dwFB-E2H2K64g9OfERNUQcfP0eb/view?usp=drive_link)
5. Power and Pre-treatment Fit: Optimizing Synthetic Control Method for Quasi-experiments, Ali O Polat (Shipt Inc.)*; [Slides](https://drive.google.com/file/d/1F9bCmoCYg7AeNxKXekCze7a4ELXs2Ohv/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1rDsCmwl23HELiD-P_Qq9TUZZ_3opbEji/view?usp=drive_link)
6. Dynamic Causal Structure Discovery and Causal Effect Estimation, Jianian Wang (North Carolina State Unicersity)*; Rui Song (North Carolina State Unicersity); [Slides](https://drive.google.com/file/d/13MzjyZLGxfGNoAg3TMRF0Es-TqtnRVSH/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1928QcX3PdFan_gkeJl-mxEgl7kQ9KupN/view?usp=drive_link)
7. Hierarchical Clustering As a Novel Solution to the Notorious Multicollinearity Problem in Observational Causal Inference, Yufei Wu (Airbnb)*; Zhiying Gu (Airbnb); Alex Deng (Airbnb); Jacob Zhu (Airbnb); [Slides](https://drive.google.com/file/d/1r0xXFFflSVDFNtGEypUGh845CeycF51q/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1HQekrFF1vNNrO7TF4aX43850-Dc6UHgo/view?usp=drive_link)
### For Poster Presentation
8. Community Detection-Enhanced Causal Structural Learning, Yuhe Gao (North Carolina State University)*; Hengrui Cai (University of California Irvine); Sheng Zhang (North Carolina State University); Rui Song (North Carolina State University); [Poster](https://drive.google.com/file/d/1dC_WLUhOJletn-kDJd1GPUJDyHwpMblG/view?usp=drive_link), [Paper](https://drive.google.com/file/d/11vnSFYDQZ1ZjztaM_eW7J9jd5tKH7y6Y/view?usp=drive_link)
9. ACE: Active Learning for Causal Inference with Expensive Experiments, Difan Song (Georgia Institute of Technology)*; Simon Mak (Duke University); C.F. Jeff Wu (Georgia Institute of Technology); [Paper](https://drive.google.com/file/d/1-fCLcT4RYAHJlsSBUuuAwsuqHDf9B4Qt/view?usp=drive_link)
10. Evaluate the Impact of Similar Products Ad Group Recommendations with Causal Inference, Jamie Chen (Amazon)*; Zuqi Shang (AmaOn); Raif Rustamov (Amazon); [Poster](https://drive.google.com/file/d/13UTLTuJKBA5hIFZ0g1HolLHQpAXhqcX4/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1ahjN_fNvDjmdwD0btW4UvgxaKgCB32kP/view?usp=drive_link)
11. Machine Learning based Framework for Robust Price-Sensitivity Estimation with Application to Airline Pricing, Shahin Boluki (Pros Inc)*; Ravi Kumar (PROS); [Poster](https://drive.google.com/file/d/1AcRtI5NtovYWMEOLnQ7dE_IRgyHVukqf/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1n5mL1quGS5jWVZXXqUrL349CJH-I2BOW/view?usp=drive_link)
12. OpportunityFinder: A Framework for Automated Causal Inference, Huy Nguyen (Amazon)*; Prince Grover (Amazon); Devashish Khatwani (Amazon); [Poster](https://drive.google.com/file/d/1MH_vV5MDafqAOVLnQRIgmG1IFGXZx05r/view?usp=drive_link), [Paper](https://drive.google.com/file/d/1_yAoohM0jG0uPi7om9_6igVQJI1JM-Ce/view?usp=drive_link)
## **Organizers**
* Chu Wang, Amazon
* Yingfei Wang, University of Washington
* Xinwei Ma, UC San Diego
* [Zeyu Zheng](mailto:zyzheng@berkeley.edu), UC Berkeley, Amazon - main contact
### [CausalML](https://github.com/uber/causalml) Team
* Jing Pan, Snap, CausalML
* Yifeng Wu, Uber, CausalML
* Huigang Chen, Meta, CausalML
* Totte Harinen, AirBnB, CausalML
* Paul Lo, Snap, CausalML
* [Jeong-Yoon Lee](mailto:jeong@uber.com), Uber, CausalML - main contact
* Zhenyu Zhao, Tencent, CausalML
### [EconML](https://github.com/py-why/EconML) Team
* Fabio Vera, Microsoft Research, EconML
* Eleanor Dillon, Microsoft Research, EconML
* Keith Battocchi, Microsoft Research, EconML