We investigate the finite-time analysis of finding ($\delta,\epsilon$)-stationary points for nonsmooth nonconvex objectives in decentralized stochastic optimization. A set of agents aim at minimizing a global function using only their local information by interacting over a network. We present a novel algorithm, called Multi Epoch Decentralized Online Learning (ME-DOL), for which we establish the sample complexity in various settings. First, using a recently proposed online-to-nonconvex technique, we show that our algorithm recovers the optimal convergence rate of smooth nonconvex objectives. We then extend our analysis to the nonsmooth setting, building on properties of randomized smoothing and Goldstein-subdifferential sets. We establish the sample complexity of $O(\delta^{-1}\epsilon^{-3})$, which to the best of our knowledge is the first finite-time guarantee for decentralized nonsmooth nonconvex stochastic optimization in the first-order setting (without weak-convexity), matching its optimal centralized counterpart. We further prove the same rate for the zero-order oracle setting without using variance reduction.

We demonstrate a multi-fidelity (MF) machine learning ensemble framework for the inverse design of photonic surfaces, trained on a dataset of 11,759 samples that we fabricate using high throughput femtosecond laser processing. The MF ensemble combines an initial low fidelity model for generating design solutions, with a high fidelity model that refines these solutions through local optimization. The combined MF ensemble can generate multiple disparate sets of laser-processing parameters that can each produce the same target input spectral emissivity with high accuracy (root mean squared errors < 2%). SHapley Additive exPlanations analysis shows transparent model interpretability of the complex relationship between laser parameters and spectral emissivity. Finally, the MF ensemble is experimentally validated by fabricating and evaluating photonic surface designs that it generates for improved efficiency energy harvesting devices. Our approach provides a powerful tool for advancing the inverse design of photonic surfaces in energy harvesting applications.

Deep learning models can extract predictive and actionable information from complex inputs. The richer the inputs, the better these models usually perform. However, models that leverage rich inputs (e.g., multi-modality) can be difficult to deploy widely, because some inputs may be missing at inference. Current popular solutions to this problem include marginalization, imputation, and training multiple models. Marginalization can obtain calibrated predictions but it is computationally costly and therefore only feasible for low dimensional inputs. Imputation may result in inaccurate predictions because it employs point estimates for missing variables and does not work well for high dimensional inputs (e.g., images). Training multiple models whereby each model takes different subsets of inputs can work well but requires knowing missing input patterns in advance. Furthermore, training and retaining multiple models can be costly. We propose an efficient way to learn both the conditional distribution using full inputs and the marginal distributions. Our method, Knockout, randomly replaces input features with appropriate placeholder values during training. We provide a theoretical justification of Knockout and show that it can be viewed as an implicit marginalization strategy. We evaluate Knockout in a wide range of simulations and real-world datasets and show that it can offer strong empirical performance.

Mixup and its variants form a popular class of data augmentation techniques.Using a random sample pair, it generates a new sample by linear interpolation of the inputs and labels. However, generating only one single interpolation may limit its augmentation ability. In this paper, we propose a simple yet effective extension called multi-mix, which generates multiple interpolations from a sample pair. With an ordered sequence of generated samples, multi-mix can better guide the training process than standard mixup. Moreover, theoretically, this can also reduce the stochastic gradient variance. Extensive experiments on a number of synthetic and large-scale data sets demonstrate that multi-mix outperforms various mixup variants and non-mixup-based baselines in terms of generalization, robustness, and calibration.

The integration of Differential Privacy (DP) with diffusion models (DMs) presents a promising yet challenging frontier, particularly due to the substantial memorization capabilities of DMs that pose significant privacy risks. Differential privacy offers a rigorous framework for safeguarding individual data points during model training, with Differential Privacy Stochastic Gradient Descent (DP-SGD) being a prominent implementation. Diffusion method decomposes image generation into iterative steps, theoretically aligning well with DP's incremental noise addition. Despite the natural fit, the unique architecture of DMs necessitates tailored approaches to effectively balance privacy-utility trade-off. Recent developments in this field have highlighted the potential for generating high-quality synthetic data by pre-training on public data (i.e., ImageNet) and fine-tuning on private data, however, there is a pronounced gap in research on optimizing the trade-offs involved in DP settings, particularly concerning parameter efficiency and model scalability. Our work addresses this by proposing a parameter-efficient fine-tuning strategy optimized for private diffusion models, which minimizes the number of trainable parameters to enhance the privacy-utility trade-off. We empirically demonstrate that our method achieves state-of-the-art performance in DP synthesis, significantly surpassing previous benchmarks on widely studied datasets (e.g., with only 0.47M trainable parameters, achieving a more than 35% improvement over the previous state-of-the-art with a small privacy budget on the CelebA-64 dataset). Anonymous codes available at https://anonymous.4open.science/r/DP-LORA-F02F.

Selecting suitable data for training machine learning models is crucial since large, web-scraped, real datasets contain noisy artifacts that affect the quality and relevance of individual data points. These artifacts will impact the performance and generalization of the model. We formulate this problem as a data valuation task, assigning a value to data points in the training set according to how similar or dissimilar they are to a clean and curated validation set. Recently, LAVA (Just et al., 2023) successfully demonstrated the use of optimal transport (OT) between a large noisy training dataset and a clean validation set, to value training data efficiently, without the dependency on model performance. However, the LAVA algorithm requires the whole dataset as an input, this limits its application to large datasets. Inspired by the scalability of stochastic (gradient) approaches which carry out computations on batches of data points instead of the entire dataset, we analogously propose SAVA, a scalable variant of LAVA with its computation on batches of data points. Intuitively, SAVA follows the same scheme as LAVA which leverages the hierarchically defined OT for data valuation. However, while LAVA processes the whole dataset, SAVA divides the dataset into batches of data points, and carries out the OT problem computation on those batches. We perform extensive experiments, to demonstrate that SAVA can scale to large datasets with millions of data points and doesn't trade off data valuation performance.

Differentially private empirical risk minimization (DP-ERM) is a fundamental problem in private optimization. While the theory of DP-ERM is well-studied, as large-scale models become prevalent, traditional DP-ERM methods face new challenges, including (1) the prohibitive dependence on the ambient dimension, (2) the highly non-smooth objective functions, (3) costly first-order gradient oracles. Such challenges demand rethinking existing DP-ERM methodologies. In this work, we show that the regularized exponential mechanism combined with existing samplers can address these challenges altogether: under the standard unconstrained domain and low-rank gradients assumptions, our algorithm can achieve rank-dependent risk bounds for non-smooth convex objectives using only zeroth order oracles, which was not accomplished by prior methods. This highlights the power of sampling in differential privacy. We further construct lower bounds, demonstrating that when gradients are full-rank, there is no separation between the constrained and unconstrained settings. Our lower bound is derived from a general black-box reduction from unconstrained to the constrained domain and an improved lower bound in the constrained setting, which might be of independent interest.

Moreover, due to the data-driven nature of digital platforms, interesting dynamics emerge among users and service In digital markets comprised of many competing providers: on the one hand, users choose amongst services, each user chooses between multiple providers based on the quality of their services; on the other service providers according to their preferences, hand, providers use the user data to improve and update and the chosen service makes use of the user data their services, affecting future user choices (Ginart et al., to incrementally improve its model. The service 2021; Kwon et al., 2022; Dean et al., 2024; Jagadeesan et al., providers' models influence which service the 2023a). For example, in personalized music streaming platform, user will choose at the next time step, and the a user chooses amongst different music streaming user's choice, in return, influences the model update, platforms based on how well they meet the user's needs.

Machine unlearning strives to uphold the data owners' right to be forgotten by enabling models to selectively forget specific data. Recent advances suggest precomputing and storing statistics extracted from second-order information and implementing unlearning through Newton-style updates. However, the theoretical analysis of these works often depends on restrictive assumptions of convexity and smoothness, and those mentioned operations on Hessian matrix are extremely costly. As a result, applying these works to high-dimensional models becomes challenging. In this paper, we propose an efficient Hessian-free certified unlearning. We propose to maintain a statistical vector for each data, computed through affine stochastic recursion approximation of the difference between retrained and learned models. Our analysis does not involve inverting Hessian and thus can be extended to non-convex non-smooth objectives. Under same assumptions, we demonstrate advancements of proposed method beyond the state-of-the-art theoretical studies, in terms of generalization, unlearning guarantee, deletion capacity, and computation/storage complexity, and we show that the unlearned model of our proposed approach is close to or same as the retrained model. Based on the strategy of recollecting statistics for forgetting data, we develop an algorithm that achieves near-instantaneous unlearning as it only requires a vector addition operation. Experiments demonstrate that the proposed scheme surpasses existing results by orders of magnitude in terms of time/storage costs, while also enhancing accuracy.

Abstract--Federated learning (FL) has emerged as a collaborative approach that allows multiple clients to jointly learn a machine learning model without sharing their private data. The concern about privacy leakage, albeit demonstrated under specific conditions [1], has triggered numerous follow-up research in designing powerful attacking methods and effective defending mechanisms aiming to thwart these attacking methods. Nevertheless, privacy-preserving mechanisms employed in these defending methods invariably lead to compromised model performances due to a fixed obfuscation applied to private data or gradients. In this article, we, therefore, propose a novel adaptive obfuscation mechanism, coined FedAdOb, to protect private data without yielding original model performances. T echnically, FedAdOb utilizes passport-based adaptive obfuscation to ensure data privacy in both horizontal and vertical federated learning settings. The privacy-preserving capabilities of FedAdOb, specifically with regard to private features and labels, are theoretically proven through Theorems 1 and 2. Furthermore, extensive experimental evaluations conducted on various datasets and network architectures demonstrate the effectiveness of FedAdOb by manifesting its superior trade-off between privacy preservation and model performance, surpassing existing methods. Federated Learning (FL) offers a privacy-preserving framework that allows multiple organizations to jointly build global models without disclosing private datasets [2], [3], [4], [5]. Two distinct paradigms have been proposed in the context of FL [5]: Horizontal Federated Learning (HFL) and V ertical Federated Learning (VFL). HFL focuses on scenarios where multiple entities have similar features but different samples. It is suitable for cases where data sources are distributed, such as healthcare institutions contributing patient data for disease prediction. On the other hand, VFL addresses situations where entities hold different attributes or features of the same samples. This approach is useful in scenarios like combining demographic information from banks with call records from telecom companies to predict customer behavior. Since the introduction of HFL and VFL, studies have highlighted the existence of privacy risks in specific scenarios.