Data

Tools

Indexes

Applications

Experiments

Awesome

An open API service indexing awesome lists of open source software.

https://github.com/bertcarnell/jsm2024


https://github.com/bertcarnell/jsm2024

Last synced: 16 days ago
JSON representation

Awesome Lists containing this project

README 

        
# jsm2024



The Joint Statistical Meetings (JSM) were held in Portland, Oregon Aug 3-8, 2024.



![JSM Logo](img/jsm_2024_logo.png)



## NISS Luncheon (8/4 1200)



### Veridical Data Science



Talked with Professor Bin Yu (UC Berkeley) about her book: [Veridical Data Science](https://vdsbook.com/).



### AI in Practice:  Transitions and Tools



Dr. Nancy McMillan, Battelle



### Harvard Data Science Review



Xiao-Li Meng showed this great resource [here](https://hdsr.mitpress.mit.edu/) and [here](https://datascience.harvard.edu/)



## Recent Developments in Causal Inference (8/4 1600)



### A Meta-Learning Method fo Estimation of Causal Excursion Effects to Assess Time-Varying Moderation



Micro-randomized Trials - The idea that you can treat a longitudinal study where the same treatment is applied multiple times and treat it like a raondomized trial.



Applications to Next Best Offer are possible.



Jieru Shi - University of Michigan



Need to learn more about doubly robust estimators for causal inference:  They estimate the Averate Treatment Effect using both regression and inverse probability weighting methods.



### Other Presentation



Beeswarm plots [here](https://r-graph-gallery.com/beeswarm.html) and [here](https://shap.readthedocs.io/en/latest/example_notebooks/api_examples/plots/beeswarm.html)



Can we add this to the torndado package?



## 2024 ASA Statistics in Marketing Doctoral Research Award Finalists Presentation (8/4 0830)



### A Scalable Recommendation Engine for New Users and Items - Boya Xu - Duke



### Optimal Comprehensible Targeting



Method to make decision tree-type comprehensible targeting narratives.  - Walter Zhang



Price Elasticity



### Real Time Personalization in Dynamic Environments - Hong Deng



### Selecting Data and Parameter Granularities:  A Bayesian Dual-Network Clustering Approach - Mingyung Kim - Wharton School



### Targeted Marketing with Large Batches - Keyan Li - MIT



## Bayesian Causal Inference (8/5 1030)



### A Semiparametric Bayesian Approach for Extreme Quantile Treatment Effects for Heavy-tailed data - Arnab  Aich



Could be useful for balances



### Bayesian Causal Prediction:  Multivariate Graphical Dynamic Models - Luke Vrotsos - Duke



Great talk on how to do Bayesian Synthetic controls.



With Mike West at Duke



references:



- [Comparative Politics and the Synthetic Control](https://economics.mit.edu/sites/default/files/publications/Comparative%20Politics%20and%20the%20Synthetic%20Control.pdf)

- [The Economic Costs of Conflict: A Case Study of the Basque Country](https://economics.mit.edu/sites/default/files/publications/The%20Economic%20Costs%20of%20Conflict.pdf)

- [Inferring causal impact using Bayesian structural time-series models](https://arxiv.org/abs/1506.00356)

- [GPU Accelerated Bayesian Learning and Forcasting in Simultaneous Graphical Dynamic Linear Models](https://projecteuclid.org/journals/bayesian-analysis/volume-11/issue-1/GPU-Accelerated-Bayesian-Learning-and-Forecasting-in-Simultaneous-Graphical-Dynamic/10.1214/15-BA946.pdf)

    - [code](https://github.com/lutzgruber/gpuSGDLM)



### Bayesian Sensitivity Analysis for Extending Inferences to a Target Population without Positivity - Jun Lu - University Illinois



### Causally Sound Priors for Binary Experiments - Nicholas Irons - U Washington



### Horeshoe Priors for Sparse Dirichlet-Multinomial Models - Yuexi Want - University of Illinois



### Incorporate external control data into RCT using propensity score stratification and mixture prior - Xun Xu - U Texas



## Advanced Statistical Methods in Non-parametric Statistics and Causal Inference for Complex Data Structures (8/5 1400)



### Identifying Significant Mediators in High Dimensional Causal Graphical Models with FDR Control - Xiufan Yu, Notre Dame



### De-confounding Causal Inference using Latent Multiple-Mediator Pathways - Yubai Yuan - Penn State



### Causal Inference on Distribution Functions - Dehan Kong - U Toronto



[paper](https://arxiv.org/abs/2101.01599)

[code](https://github.com/kongdehanstat/causaldistributionfunction)



### Nonlinear Global Frechet Regression for Random Objects via Weak Conditional Expectation - Satarupa Bhattacharjee - Penn State



### Smoothed Robust Phase Retrieval - Zhong Zheng - Penn State



## The Application of Generative Models in Marketing Research and Practice (8/6 1030)



### Using GPT for Market Research - Ayelet Israeli - Harvard Business School



- Used GPT as a random customer sampling mechanism and examined the distribution of its responses

- When prompted as a random customer, GPT exhibits behaviors consistent with economic theory (price elasticity)

- GPT-based estimates are realistic and consistent with values obtained from existing research



### Using Multimodal LLM to Extract and Discover Features from Ad Images - Poppy Zhang, Meta



Trained a classifier to extrat the important parts of an image for Meta customers



[nielsen study](https://www.nielsen.com/insights/2017/when-it-comes-to-advertising-effectiveness-what-is-key/)



Lots of other Insights from Nielsen

 - [here](https://www.nielsen.com/insights/)

 - [here](https://www.nielsen.com/insights/2022/roi-report/)

 - [here](https://www.nielsen.com/insights/2024/are-you-investing-performance-marketing-for-right-reasons/)



Multi-model LLM



[LLAVA](https://llava-vl.github.io/)



Results:  Top Advertisers (controlling for sales region, industry, and campaign size)



- show price (negative)

- contain promotion

- show positive emotion

- show experience of using the product

- comparative

- show endorsement

- use humor

- visually complext (negative)



## IMS Medallion Lecture:  Winners with Confidence:  Discrete Argmin Inference Using Cross-Validated Exponential Mechanism - Ji Zhu (8/6 1400)



[paper](https://arxiv.org/abs/2408.02060)



Idea was how to say which machine learning results were actually significantly different from each other



## Robust Causal Inference (8/7 0830)



### Adaptive Experiments with Delayed Outcomes - Waverly Wei - USC



Create adaptive experiments where you measure some things early, and then other measures later, and you want to adaptively change the experiment



### Bias Correction for Randomization-Based Average Treatment Effect Estimation in Inexactly Matched Observational Studies - Siyu Heng - NYU



### Design-based causal inference for balanced incomplete block designs - Nicole Pashley - Rutgers



[Nicole Pashley](https://sites.google.com/view/npashley/home?authuser=0)

[Taehyeon Koo](https://taehyeonkoo.github.io/)



### Random distribution shift and causal inference - Zhonghua Liu - Columbia



## New Methods in Causal Inference and Reinforcement Learning for Personalized Decision-Making (8/7 1030)



### The promises of multiple outcomes - Linbo Wang - U Toronto



[website](https://sites.google.com/site/linbowangpku/home)

[paper](https://arxiv.org/abs/2012.05849)

[slides](https://drive.google.com/file/d/1jwkefODWm7KBSzl0RNMUT9MABIC8YbIE/view)



### Balancing Personalization and Pooling: Decision-making and Statistical Inference with Limited Time Horizons - Yongyi Guo



### Did We Personalize? Assessing Personalization by an Online Reinforcement Learning Algorithm Using Resampling - Raaz Dwivendi - UC Berkeley



### Further Results on Target Trials and Structural Nested Models: Emulating RCTs using Observational Longitudinal Data - Anish Agarwal - Columbia



## Causal Inference on Networks and Complex Data Structures (8/7 1400)



### Analyzing Political Polarization in the Presence of Partially Observed Social Networks - Sharmodeep Bhattacharyya, Oregon State University



### Causal Inference Under Network Interference with Noise - Daniel Sussman, Boston University



### Exploratory Data Analysis, Confirmatory Data Analysis and Replication in the Same Observational Study: A Two Team Cross-Screening Approach to Studying the Effect of Unwanted Pregnancy on Mothers' Later Life Outcomes - Dylan Small, University of Pennsylvania



### Nonsense Associations in Markov Random Fields - Elizabeth Ogburn, Johns Hopkins University



## COPSS Distinguished Achievement Award and Lectureship (8/7 1600)



#### Pre-Training and the Lasso



[paper](https://arxiv.org/abs/2401.12911)

[paper](https://arxiv.org/html/2401.12911v1)

[similar presentation - same slides](https://www.youtube.com/watch?v=zIGc5Z2MaYM)

[ptLasso](https://github.com/erincr/ptLasso/)

[web](https://erincr.github.io/ptLasso/)

[Erin Craig](https://www.erincraig.me/)