arXiv subjects

Shan Ba

Publications and source records attributed to Shan Ba.

8 recordsLinked to original sources

arXiv2025

Enterprise Experimentation with Hierarchical Entities

In this paper, we address the challenges in running enterprise experimentation with hierarchical entities and present the methodologies behind the implementation of the Enterprise Experimentation Platform (EEP) at LinkedIn, which plays a pivotal role in delivering an intelligent, scalable, and reliable experimentation experience to optimize performance across all LinkedIn's enterprise products. We start with an introduction to the hierarchical entity relationships of the enterprise products and how such complex entity structure poses challenges to experimentation. We then delve into the details of our solutions for EEP including taxonomy based design setup with multiple entities, analysis methodologies in the presence of hierarchical entities, and advanced variance reduction techniques, etc. Recognizing the hierarchical ramping patterns inherent in enterprise experiments, we also propose a two-level Sample Size Ratio Mismatch (SSRM) detection methodology. This approach addresses SSRM at both the randomization unit and analysis unit levels, bolstering the internal validity and trustworthiness of analysis results within EEP. In the end, we discuss implementations and examine the business impacts of EEP through practical examples.

stat.ME↗

arXiv2023

Producer-Side Experiments Based on Counterfactual Interleaving Designs for Online Recommender Systems

Recommender systems have become an integral part of online platforms, providing personalized recommendations for purchases, content consumption, and interpersonal connections. These systems consist of two sides: the producer side comprises product sellers, content creators, or service providers, etc., and the consumer side includes buyers, viewers, or customers, etc. To optimize online recommender systems, A/B tests serve as the golden standard for comparing different ranking models and evaluating their impact on both the consumers and producers. While consumer-side experiments is relatively straightforward to design and commonly employed to assess the impact of ranking changes on the behavior of consumers (buyers, viewers, etc.), designing producer-side experiments for an online recommender/ranking system is notably more intricate because producer items in the treatment and control groups need to be ranked by different models and then merged into a unified ranking to be presented to each consumer. Current design solutions in the literature are ad hoc and lacking rigorous guiding principles. In this paper, we examine limitations of these existing methods and propose the principle of consistency and principle of monotonicity for designing producer-side experiments of online recommender systems. Building upon these principles, we also present a systematic solution based on counterfactual interleaving designs to accurately measure the impacts of ranking changes on the producers (sellers, creators, etc.).

stat.ME↗

arXiv2021

Towards Optimal Variance Reduction in Online Controlled Experiments

We study optimal variance reduction solutions for count and ratio metrics in online controlled experiments. Our methods leverage flexible machine learning tools to incorporate covariates that are independent from the treatment but have predictive power for the outcomes, and employ the cross-fitting technique to remove the bias in complex machine learning models. We establish CLT-type asymptotic inference based on our estimators under mild convergence conditions. Our procedures are optimal (efficient) for the corresponding targets as long as the machine learning estimators are consistent, without any requirement for their convergence rates. In complement to the general optimal procedure, we also derive a linear adjustment method for ratio metrics as a special case that is computationally efficient and can flexibly incorporate any pre-treatment covariates. We evaluate the proposed variance reduction procedures with comprehensive simulation studies and provide practical suggestions regarding commonly adopted assumptions in computing ratio metrics. When tested on real online experiment data from LinkedIn, the proposed optimal procedure for ratio metrics can reduce up to 80\% of variance compared to the standard difference-in-mean estimator and also further reduce up to 30\% of variance compared to the CUPED approach by going beyond linearity and incorporating a large number of extra covariates.

stat.ME↗

arXiv2021

Online Experimentation with Surrogate Metrics: Guidelines and a Case Study

A/B tests have been widely adopted across industries as the golden rule that guides decision making. However, the long-term true north metrics we ultimately want to drive through A/B test may take a long time to mature. In these situations, a surrogate metric which predicts the long-term metric is often used instead to conclude whether the treatment is effective. However, because the surrogate rarely predicts the true north perfectly, a regular A/B test based on surrogate metrics tends to have high false positive rate and the treatment variant deemed favorable from the test may not be the winning one. In this paper, we discuss how to adjust the A/B testing comparison to ensure experiment results are trustworthy. We also provide practical guidelines on the choice of good surrogate metrics. To provide a concrete example of how to leverage surrogate metrics for fast decision making, we present a case study on developing and evaluating the predicted confirmed hire surrogate metric in LinkedIn job marketplace.

stat.AP↗

arXiv2021

Deep Multistage Multi-Task Learning for Quality Prediction of Multistage Manufacturing Systems

In multistage manufacturing systems, modeling multiple quality indices based on the process sensing variables is important. However, the classic modeling technique predicts each quality variable one at a time, which fails to consider the correlation within or between stages. We propose a deep multistage multi-task learning framework to jointly predict all output sensing variables in a unified end-to-end learning framework according to the sequential system architecture in the MMS. Our numerical studies and real case study have shown that the new model has a superior performance compared to many benchmark methods as well as great interpretability through developed variable selection techniques.

cs.LG↗

arXiv2020

Robust Experimental Designs for Model Calibration

A computer model can be used for predicting an output only after specifying the values of some unknown physical constants known as calibration parameters. The unknown calibration parameters can be estimated from real data by conducting physical experiments. This paper presents an approach to optimally design such a physical experiment. The problem of optimally designing physical experiment, using a computer model, is similar to the problem of finding optimal design for fitting nonlinear models. However, the problem is more challenging than the existing work on nonlinear optimal design because of the possibility of model discrepancy, that is, the computer model may not be an accurate representation of the true underlying model. Therefore, we propose an optimal design approach that is robust to potential model discrepancies. We show that our designs are better than the commonly used physical experimental designs that do not make use of the information contained in the computer model and other nonlinear optimal designs that ignore potential model discrepancies. We illustrate our approach using a toy example and a real example from industry.

stat.ME↗

arXiv2017

Space-Filling Designs for Robustness Experiments

To identify the robust settings of the control factors, it is very important to understand how they interact with the noise factors. In this article, we propose space-filling designs for computer experiments that are more capable of accurately estimating the control-by-noise interactions. Moreover, the existing space-filling designs focus on uniformly distributing the points in the design space, which are not suitable for noise factors because they usually follow non-uniform distributions such as normal distribution. This would suggest placing more points in the regions with high probability mass. However, noise factors also tend to have a smooth relationship with the response and therefore, placing more points towards the tails of the distribution is also useful for accurately estimating the relationship. These two opposing effects make the experimental design methodology a challenging problem. We propose optimal and computationally efficient solutions to this problem and demonstrate their advantages using simulated examples and a real industry example involving a manufacturing packing line.

stat.ME↗

arXiv2013

Composite Gaussian process models for emulating expensive functions

A new type of nonstationary Gaussian process model is developed for approximating computationally expensive functions. The new model is a composite of two Gaussian processes, where the first one captures the smooth global trend and the second one models local details. The new predictor also incorporates a flexible variance model, which makes it more capable of approximating surfaces with varying volatility. Compared to the commonly used stationary Gaussian process model, the new predictor is numerically more stable and can more accurately approximate complex surfaces when the experimental design is sparse. In addition, the new model can also improve the prediction intervals by quantifying the change of local variability associated with the response. Advantages of the new predictor are demonstrated using several examples.

stat.AP↗