Tucker Balch

AAAI Conference 2025 Conference Paper

Auditing and Enforcing Conditional Fairness via Optimal Transport

• Mohsen Ghassemi
• Alan Mishler
• Niccolo Dalmasso
• Luhao Zhang
• Vamsi K. Potluru
• Tucker Balch
• Manuela Veloso

Conditional demographic parity (CDP) is a measure of the demographic parity of a predictive model or decision process when conditioning on an additional feature or set of features. Many algorithmic fairness techniques exist to target demographic parity, but CDP is much harder to achieve, particularly when the conditioning variable has many levels and/or when the model outputs are continuous. The problem of auditing and enforcing CDP is understudied in the literature. In light of this, we propose novel measures of conditional demographic disparity (CDD) which rely on statistical distances borrowed from the optimal transport literature. We further design and evaluate regularization-based approaches based on these CDD measures. Our methods, FairBiT and FairLeap, allow us to target conditional demographic parity even when the conditioning variable has many levels. When model outputs are continuous, our methods target full equality of the conditional distributions, unlike other methods that only consider first moments or related proxy quantities. We validate our approaches on real-world datasets.

TMLR Journal 2025 Journal Article

Variational Neural Stochastic Differential Equations with Change Points

• Yousef El-Laham
• Zhongchang Sun
• Haibei Zhu
• Tucker Balch
• Svitlana Vyetrenko

In this work, we explore modeling change points in time-series data using neural stochastic differential equations (neural SDEs). We propose a novel model formulation and training procedure based on the variational autoencoder (VAE) framework for modeling time-series as a neural SDE. Unlike existing algorithms training neural SDEs as VAEs, our proposed algorithm only necessitates a Gaussian prior of the initial state of the latent stochastic process, rather than a Wiener process prior on the entire latent stochastic process. We develop two methodologies for modeling and estimating change points in time-series data with distribution shifts. Our iterative algorithm alternates between updating neural SDE parameters and updating the change points based on either a maximum likelihood-based approach or a change point detection algorithm using the sequential likelihood ratio test. We also discuss theoretical implications of the proposed change point detection scheme. Finally, we present an empirical evaluation that demonstrates the expressive power of our proposed model, showing that it can effectively model both classical parametric SDEs and some real datasets with distribution shifts.

AAMAS Conference 2024 Conference Paper

Atlas-X Equity Financing: Unlocking New Methods to Securely Obfuscate Axe Inventory Data Based on Differential Privacy

• Antigoni Polychroniadou
• Gabriele Ciprianni
• Richard Hua
• Tucker Balch

Banks publish daily a list of available securities/assets (axe list) to selected clients to help them effectively locate Long (buy) or Short (sell) trades at reduced financing rates. This reduces costs for the bank, as the list aggregates the bank’s internal firm inventory per asset for all clients of long as well as short trades. However, this is somewhat problematic: (1) the bank’s inventory is revealed; (2) trades of clients who contribute to the aggregated list, particularly those deemed large, are revealed to other clients. Clients conducting sizable trades with the bank and possessing a portion of the aggregated asset exceeding 50% are considered to be concentrated clients. This could potentially reveal a trading concentrated client’s activity to their competitors, thus providing an unfair advantage over the market. Atlas-X Axe Obfuscation, powered by new differential private methods, enables a bank to obfuscate its published axe list on a daily basis while under continual observation, thus maintaining an acceptable inventory Profit and Loss (P&L) cost pertaining to the noisy obfuscated axe list while reducing the clients’ trading activity leakage. Our main differential private innovation is a differential private aggregator for streams (time series data) of both positive and negative integers under continual observation. For the last two years, the Atlas-X system has been live in production across three major regions—USA, Europe, and Asia—at J. P. Morgan, a major financial institution. To our knowledge, it is the first differential privacy solution to be deployed in the financial sector. We also report benchmarks of our algorithm based on (anonymous) real and synthetic data to showcase the quality of our obfuscation and its success in production.

NeurIPS Conference 2024 Conference Paper

Fair Wasserstein Coresets

• Zikai Xiong
• Niccolò Dalmasso
• Shubham Sharma
• Freddy Lecue
• Daniele Magazzeni
• Vamsi K. Potluru
• Tucker Balch
• Manuela Veloso

Data distillation and coresets have emerged as popular approaches to generate a smaller representative set of samples for downstream learning tasks to handle large-scale datasets. At the same time, machine learning is being increasingly applied to decision-making processes at a societal level, making it imperative for modelers to address inherent biases towards subgroups present in the data. While current approaches focus on creating fair synthetic representative samples by optimizing local properties relative to the original samples, their impact on downstream learning processes has yet to be explored. In this work, we present fair Wasserstein coresets ($\texttt{FWC}$), a novel coreset approach which generates fair synthetic representative samples along with sample-level weights to be used in downstream learning tasks. $\texttt{FWC}$ uses an efficient majority minimization algorithm to minimize the Wasserstein distance between the original dataset and the weighted synthetic samples while enforcing demographic parity. We show that an unconstrained version of $\texttt{FWC}$ is equivalent to Lloyd's algorithm for k-medians and k-means clustering. Experiments conducted on both synthetic and real datasets show that $\texttt{FWC}$: (i) achieves a competitive fairness-performance tradeoff in downstream models compared to existing approaches, (ii) improves downstream fairness when added to the existing training data and (iii) can be used to reduce biases in predictions from large language models (GPT-3. 5 and GPT-4).

AAAI Conference 2024 Conference Paper

FairWASP: Fast and Optimal Fair Wasserstein Pre-processing

• Zikai Xiong
• Niccolò Dalmasso
• Alan Mishler
• Vamsi K. Potluru
• Tucker Balch
• Manuela Veloso

Recent years have seen a surge of machine learning approaches aimed at reducing disparities in model outputs across different subgroups. In many settings, training data may be used in multiple downstream applications by different users, which means it may be most effective to intervene on the training data itself. In this work, we present FairWASP, a novel pre-processing approach designed to reduce disparities in classification datasets without modifying the original data. FairWASP returns sample-level weights such that the reweighted dataset minimizes the Wasserstein distance to the original dataset while satisfying (an empirical version of) demographic parity, a popular fairness criterion. We show theoretically that integer weights are optimal, which means our method can be equivalently understood as duplicating or eliminating samples. FairWASP can therefore be used to construct datasets which can be fed into any classification method, not just methods which accept sample weights. Our work is based on reformulating the pre-processing task as a large-scale mixed-integer program (MIP), for which we propose a highly efficient algorithm based on the cutting plane method. Experiments demonstrate that our proposed optimization algorithm significantly outperforms state-of-the-art commercial solvers in solving both the MIP and its linear program relaxation. Further experiments highlight the competitive performance of FairWASP in reducing disparities while preserving accuracy in downstream classification settings.

TMLR Journal 2023 Journal Article

Privacy-Preserving Energy-Based Generative Models for Marginal Distribution Protection

• Robert E. Tillman
• Tucker Balch
• Manuela Veloso

We consider learning generative models for sensitive financial and healthcare data. While previous work incorporates Differential Privacy (DP) into GAN training to protect the privacy of individual training instances, we consider a different privacy context where the primary objective is protecting the privacy of sensitive marginal distributions of the true generative process. We propose and motivate a new notion of privacy: \emph{$\alpha$-Level Marginal Distribution Privacy} ($\alpha$-LMDP), which provides a statistical guarantee that the sensitive generative marginal distributions are different from the observed real data. We then propose \emph{Privacy-Preserving Energy Models (PPEMs)}, a novel energy-based generative model formulation where the representations for these attributes are isolated from other attributes. This structured formulation motivates a learning procedure where a penalty based on a statistical goodness of fit test, the \emph{Kernel Stein Discrepancy}, can be applied to only the attributes requiring privacy so that $\alpha$-LMDP may be satisfied without affecting the other attributes. We evaluate this approach using financial and healthcare datasets and demonstrate that the resulting learnt generative models produce high fidelity synthetic data while preserving privacy. We also show that PPEMs can incorporate both $\alpha$-LMDP \emph{and} DP in contexts where both forms of privacy are required.

AAAI Conference 2017 Conference Paper

Sampling Beats Fixed Estimate Predictors for Cloning Stochastic Behavior in Multiagent Systems

• Brian Hrolenok
• Byron Boots
• Tucker Balch

Modeling stochastic multiagent behavior such as fish schooling is challenging for fixed-estimate prediction techniques because they fail to reliably reproduce the stochastic aspects of the agents behavior. We show how standard fixed-estimate predictors fit within a probabilistic framework, and suggest the reason they work for certain classes of behaviors and not others. We quantify the degree of mismatch and offer alternative sampling-based modeling techniques. We are specifically interested in building executable models (as opposed to statistical or descriptive models) because we want to reproduce and study multiagent behavior in simulation. Such models can be used by biologists, sociologists, and economists to explain and predict individual and group behavior in novel scenarios, and to test hypotheses regarding group behavior. Developing models from observation of real systems is an obvious application of machine learning. Learning directly from data eliminates expensive hand processing and tuning, but introduces unique challenges that violate certain assumptions common in standard machine learning approaches. Our framework suggests a new class of sampling-based methods, which we implement and apply to simulated deterministic and stochastic schooling behaviors, as well as the observed schooling behavior of real fish. Experimental results show that our implementation performs comparably with standard learning techniques for deterministic behaviors, and better on stochastic behaviors.

AAMAS Conference 2006 Conference Paper

Learning Executable Agent Behaviors from Observation

• Andrew Guillory
• Hai Nguyen
• Tucker Balch
• Charles Isbell

AAAI Conference 2005 Conference Paper

Data-Driven MCMC for Learning and Inference in Switching Linear Dynamic Systems

• Sangmin Oh
• Tucker Balch

Switching Linear Dynamic System (SLDS) models are a popular technique for modeling complex nonlinear dynamic systems. An SLDS has significantly more descriptive power than an HMM, but inference in SLDS models is computationally intractable. This paper describes a novel inference algorithm for SLDS models based on the Data- Driven MCMC paradigm. We describe a new proposal distribution which substantially increases the convergence speed. Comparisons to standard deterministic approximation methods demonstrate the improved accuracy of our new approach. We apply our approach to the problem of learning an SLDS model of the bee dance. Honeybees communicate the location and distance to food sources through a dance that takes place within the hive. We learn SLDS model parameters from tracking data which is automatically extracted from video. We then demonstrate the ability to successfully segment novel bee dances into their constituent parts, effectively decoding the dance of the bees.