Author name cluster

Yan Dai

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

6 papers

1 author row

EAAI Journal 2026 Journal Article

A zero-shot tree-structured multi-objective evolutionary Neural Architecture Search

Yan Dai
Lixin Wei
Ziyu Hu
Hao Sun
Qianao Xu
Kexin Zhang
Boya Zhao

Neural Architecture Search (NAS) enables the automated design of high-performance neural networks; however, its practical application is often constrained by substantial computational costs, the limited reliability of single-objective proxy metrics, and insufficient modeling of architectural information flow. To address these limitations, we propose Tree-structured Evolutionary Neural Architecture Search (TreeNAS). The method integrates three components: (1) a tree-structured encoding with refinement to preserve backbone information paths under mutation; (2) a zero-cost multi-objective evaluation that jointly assesses trainability, generalization, and complexity, thereby mitigating instability from single-objective proxy; and (3) a Pareto-dominance-guided evolutionary search to encourage diverse, balanced architectures across objectives. On the standard Neural Architecture Search Benchmark 101 (NAS-Bench-101) and Neural Architecture Search Benchmark 201 (NAS-Bench-201) datasets, TreeNAS achieves state-of-the-art accuracy with a 40 × reduction in search cost. On the ImageNet dataset under strict floating-point operation (FLOPs) budgets, TreeNAS achieves accuracy comparable to training-based NAS methods while keeping the search cost to 0. 45 Graphics Processing Unit (GPU) days. Additionally, TreeNAS generalizes across modalities, from two-dimensional images to medical signals and volumetric imaging, demonstrating its potential in practical medical imaging applications.

Details DOI

NeurIPS Conference 2025 Conference Paper

Incentive-Aware Dynamic Resource Allocation under Long-Term Cost Constraints

Yan Dai
Negin Golrezaei
Patrick Jaillet

Motivated by applications such as cloud platforms allocating GPUs to users or governments deploying mobile health units across competing regions, we study the constrained dynamic allocation of a reusable resource to a group of strategic agents. Our objective is to simultaneously (i) maximize social welfare, (ii) satisfy multi-dimensional long-term cost constraints, and (iii) incentivize truthful reporting. We begin by numerically evaluating primal-dual methods widely used in constrained online optimization and find them to be highly fragile in strategic settings -- agents can easily manipulate their reports to distort future dual updates for future gain. To address this vulnerability, we develop an incentive-aware framework that makes primal-dual methods robust to strategic behavior. Our primal-side design combines epoch-based lazy updates -- discouraging agents from distorting dual updates -- with dual-adjust pricing and randomized exploration techniques that extract approximately truthful signals for learning. On the dual side, we design a novel online learning subroutine to resolve a circular dependency between actions and predictions; this makes our mechanism achieve $\tilde{\mathcal{O}}(\sqrt{T})$ social welfare regret (where $T$ is the number of allocation rounds), satisfies all cost constraints, and ensures incentive alignment. This $\tilde{\mathcal{O}}(\sqrt{T})$ performance matches that of non-strategic allocation approaches while additionally exhibiting robustness to strategic agents.

PDF Details

YNIMG Journal 2025 Journal Article

Investigating the effects of calibration errors on the spatial resolution of OPM-MEG beamformer imaging

Shengjie Qi
Xinda Song
Le Jia
Zhaoxin Duan
Yan Dai
Jing Zhang
Xiaolin Ning

The use of optically pumped magnetometers (OPMs) has provided a feasible, moveable and wearable alternative to superconducting detectors for magnetoencephalography (MEG) measurements. Recently, the widely used beamformer imaging technique has greatly improved spatial accuracy of MEG in the field of source reconstruction of neuroimaging. The spatial resolution of the source reconstruction using beamformer imaging technique was explored in the present study. The spatial accuracy of a beamformer reconstruction depends on accurate estimation of the data covariance matrix and lead field. In practical measurements, many sensor calibration errors including the gain error, crosstalk and angular error of the sensitive axis of OPMs due to for example, the low frequency magnetic field drift will distort the measured data as well as the forward model and thus reduce spatial resolution. The theory of OPM calibration errors was first provided based on the Bloch equations. The calibration errors are then quantified using the self-developed OPM array. And an analytical relationship between the Frobenius norm of the covariance matrix error and gain error, crosstalk was derived. The relationship between point-spread function (PSF) and the forward model error caused by the angular error of sensitive axis was analyzed. Finally, the effects of calibration errors on spatial resolution of OPM-MEG were investigated using simulations of two dipoles with orthogonal signals at the source level based on realistic head models. We find the presence of calibration errors will decrease the spatial resolution of beamformer reconstruction. And this decrease will become more severe as the signal-to-noise ratio increases.

Details DOI

NeurIPS Conference 2023 Conference Paper

The Crucial Role of Normalization in Sharpness-Aware Minimization

Yan Dai
Kwangjun Ahn
Suvrit Sra

Sharpness-Aware Minimization (SAM) is a recently proposed gradient-based optimizer (Foret et al. , ICLR 2021) that greatly improves the prediction performance of deep neural networks. Consequently, there has been a surge of interest in explaining its empirical success. We focus, in particular, on understanding the role played by normalization, a key component of the SAM updates. We theoretically and empirically study the effect of normalization in SAM for both convex and non-convex functions, revealing two key roles played by normalization: i) it helps in stabilizing the algorithm; and ii) it enables the algorithm to drift along a continuum (manifold) of minima -- a property identified by recent theoretical works that is the key to better performance. We further argue that these two properties of normalization make SAM robust against the choice of hyper-parameters, supporting the practicality of SAM. Our conclusions are backed by various experiments.

PDF Details

NeurIPS Conference 2022 Conference Paper

Follow-the-Perturbed-Leader for Adversarial Markov Decision Processes with Bandit Feedback

Yan Dai
Haipeng Luo
Liyu Chen

We consider regret minimization for Adversarial Markov Decision Processes (AMDPs), where the loss functions are changing over time and adversarially chosen, and the learner only observes the losses for the visited state-action pairs (i. e. , bandit feedback). While there has been a surge of studies on this problem using Online-Mirror-Descent (OMD) methods, very little is known about the Follow-the-Perturbed-Leader (FTPL) methods, which are usually computationally more efficient and also easier to implement since it only requires solving an offline planning problem. Motivated by this, we take a closer look at FTPL for learning AMDPs, starting from the standard episodic finite-horizon setting. We find some unique and intriguing difficulties in the analysis and propose a workaround to eventually show that FTPL is also able to achieve near-optimal regret bounds in this case. More importantly, we then find two significant applications: First, the analysis of FTPL turns out to be readily generalizable to delayed bandit feedback with order-optimal regret, while OMD methods exhibit extra difficulties (Jin et al. , 2022). Second, using FTPL, we also develop the first no-regret algorithm for learning communicating AMDPs in the infinite-horizon setting with bandit feedback and stochastic transitions. Our algorithm is efficient assuming access to an offline planning oracle, while even for the easier full-information setting, the only existing algorithm (Chandrasekaran and Tewari, 2021) is computationally inefficient.

PDF Details

AAAI Conference 2021 Conference Paper

RSGNet: Relation based Skeleton Graph Network for Crowded Scenes Pose Estimation

Yan Dai
Xuanhan Wang
Lianli Gao
Jingkuan Song
Heng Tao Shen

Despite of the recent great progress on multi-person pose estimation, existing solutions still remain challenging under the condition of “crowded scenes”, where RGB images capture complex real-world scenes with highly-overlapped people, severe occlusions and diverse postures. In this work, we focus on two main problems: 1) how to design an effective pipeline for crowded scenes pose estimation; and 2) how to equip this pipeline with the ability of relation modeling for interference resolving. To tackle these problems, we propose a new pipeline named Relation based Skeleton Graph Network (RSGNet). Unlike existing works that directly predict joints-of-target by labeling joints-of-interference as false positive, we first encourage all joints to be predicted. And then, a Target-aware Relation Parser (TRP) is designed to model the relation over all predicted joints, resulting in a targetaware encoding. This new pipeline will largely relieve the confusion of the joints estimation model when seeing identical joints with totally distinct labels (e. g. , the identical hand exists in two bounding boxes). Furthermore, we introduce a Skeleton Graph Machine (SGM) to model the skeletonbased commonsense knowledge, aiming to estimate the target pose with the constraint of human body structure. Such skeleton-based constraint can help to deal with the challenges in crowded scenes from a reasoning perspective. Solid experiments on pose estimation benchmarks demonstrate that our method outperforms existing state-of-the-art methods. The code and pre-trained models are publicly available online1.

PDF Details