Source author record

Min Lu

Min Lu appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Graphics Computer Science and Game Theory Computer Vision cond-mat.str-el Human-Computer Interaction Machine Learning nlin.CD Populations and Evolution quant-ph

Catalog footprint

What is connected

8works

9topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Optimal Boost Design for Auto-bidding Mechanism with Publisher Quality Constraints

Online bidding serves as a fundamental information system in mobile ecosystems, facilitating real-time ad allocation across billions of devices while optimizing both platform performance and user experience through data-driven decision making. Improving ad allocation efficiency is a long-standing research problem, as it directly enhances the economic outcomes for all participants in advertising platforms. This paper investigates the design of optimal boost factors in online bidding while incorporating quality value (the impact of displayed ads on publishers' long-term benefits). To address the divergent interests on quality, we establish a three-party auction framework with a unified welfare metric of advertiser and publisher. Within this framework, we derive the theoretical efficiency lower bound for C-competitive boost in second-price single-slot auctions, then design a novel quality-involved Boosting (q-Boost) algorithm for computing the optimal boost factor. Experimental validation on Alibaba's public dataset (AuctionNet) demonstrates 2%-6% welfare improvements over conventional approaches, proving our method's effectiveness in real-world settings.

preprint2022arXiv

HiTailor: Interactive Transformation and Visualization for Hierarchical Tabular Data

Tabular visualization techniques integrate visual representations with tabular data to avoid additional cognitive load caused by splitting users' attention. However, most of the existing studies focus on simple flat tables instead of hierarchical tables, whose complex structure limits the expressiveness of visualization results and affects users' efficiency in visualization construction. We present HiTailor, a technique for presenting and exploring hierarchical tables. HiTailor constructs an abstract model, which defines row/column headings as biclustering and hierarchical structures. Based on our abstract model, we identify three pairs of operators, Swap/Transpose, ToStacked/ToLinear, Fold/Unfold, for transformations of hierarchical tables to support users' comprehensive explorations. After transformation, users can specify a cell or block of interest in hierarchical tables as a TableUnit for visualization, and HiTailor recommends other related TableUnits according to the abstract model using different mechanisms. We demonstrate the usability of the HiTailor system through a comparative study and a case study with domain experts, showing that HiTailor can present and explore hierarchical tables from different viewpoints. HiTailor is available at https://github.com/bitvis2021/HiTailor.

preprint2021arXiv

Clusterplot: High-dimensional Cluster Visualization

We present Clusterplot, a multi-class high-dimensional data visualization tool designed to visualize cluster-level information offering an intuitive understanding of the cluster inter-relations. Our unique plots leverage 2D blobs devised to convey the geometrical and topological characteristics of clusters within the high-dimensional data, and their pairwise relations, such that general inter-cluster behavior is easily interpretable in the plot. Class identity supervision is utilized to drive the measuring of relations among clusters in high-dimension, particularly, proximity and overlap, which are then reflected spatially through the 2D blobs. We demonstrate the strength of our clusterplots and their ability to deliver a clear and intuitive informative exploration experience for high-dimensional clusters characterized by complex structure and significant overlap.

preprint2020arXiv

Dynamic Competing Risk Modeling COVID-19 in a Pandemic Scenario

The emergence of coronavirus disease 2019 (COVID-19) in the United States has forced federal and local governments to implement containment measures. Moreover, the severity of the situation has sparked engagement by both the research and clinical community with the goal of developing effective treatments for the disease. This article proposes a time dynamic prediction model with competing risks for the infected individual and develops a simple tool for policy makers to compare different strategies in terms of when to implement the strictest containment measures and how different treatments can increase or suppress infected cases. Two types of containment strategies are compared: (1) a constant containment strategy that could satisfy the needs of citizens for a long period; and (2) an adaptive containment strategy whose strict level changes across time. We consider how an effective treatment of the disease can affect the dynamics in a pandemic scenario. For illustration we consider a region with population 2.8 million and 200 initial infectious cases assuming a 4% mortality rate compared with a 2% mortality rate if a new drug is available. Our results show compared with a constant containment strategy, adaptive containment strategies shorten the outbreak length and reduce maximum daily number of cases. This, along with an effective treatment plan for the disease can minimize death rate.

preprint2020arXiv

Field-aware Calibration: A Simple and Empirically Strong Method for Reliable Probabilistic Predictions

It is often observed that the probabilistic predictions given by a machine learning model can disagree with averaged actual outcomes on specific subsets of data, which is also known as the issue of miscalibration. It is responsible for the unreliability of practical machine learning systems. For example, in online advertising, an ad can receive a click-through rate prediction of 0.1 over some population of users where its actual click rate is 0.15. In such cases, the probabilistic predictions have to be fixed before the system can be deployed. In this paper, we first introduce a new evaluation metric named field-level calibration error that measures the bias in predictions over the sensitive input field that the decision-maker concerns. We show that existing post-hoc calibration methods have limited improvements in the new field-level metric and other non-calibration metrics such as the AUC score. To this end, we propose Neural Calibration, a simple yet powerful post-hoc calibration method that learns to calibrate by making full use of the field-aware information over the validation set. We present extensive experiments on five large-scale datasets. The results showed that Neural Calibration significantly improves against uncalibrated predictions in common metrics such as the negative log-likelihood, Brier score and AUC, as well as the proposed field-level calibration error.

preprint2016arXiv

Emergent infinite-randomness fixed points from the extensive random bipartitions of the spin-1 Affleck-Kennedy-Lieb-Tasaki topological state

Quantum entanglement under an extensive bipartition can reveal the critical boundary theory of a topological phase in the parameter space. In this study we demonstrate that the infinite-randomness fixed point for spin-1/2 degrees of freedom can emerge from an extensive random bipartition of the spin-1 Affleck-Kennedy-Lieb-Tasaki chain. The nested entanglement entropy of the ground state of the reduced density matrix exhibits a logarithmic scaling with an effective central charge $\tilde{c} = 0.72 \pm 0.02 \approx \ln 2$. We further discuss, in the language of bulk quantum entanglement, how to understand all phase boundaries and the surrounding Griffiths phases for the antiferromagnetic Heisenberg spin-1 chain with quenched disorder and dimerization.

preprint2013arXiv

Rotational Projection Statistics for 3D Local Surface Description and Object Recognition

Recognizing 3D objects in the presence of noise, varying mesh resolution, occlusion and clutter is a very challenging task. This paper presents a novel method named Rotational Projection Statistics (RoPS). It has three major modules: Local Reference Frame (LRF) definition, RoPS feature description and 3D object recognition. We propose a novel technique to define the LRF by calculating the scatter matrix of all points lying on the local surface. RoPS feature descriptors are obtained by rotationally projecting the neighboring points of a feature point onto 2D planes and calculating a set of statistics (including low-order central moments and entropy) of the distribution of these projected points. Using the proposed LRF and RoPS descriptor, we present a hierarchical 3D object recognition algorithm. The performance of the proposed LRF, RoPS descriptor and object recognition algorithm was rigorously tested on a number of popular and publicly available datasets. Our proposed techniques exhibited superior performance compared to existing techniques. We also showed that our method is robust with respect to noise and varying mesh resolution. Our RoPS based algorithm achieved recognition rates of 100%, 98.9%, 95.4% and 96.0% respectively when tested on the Bologna, UWA, Queen's and Ca' Foscari Venezia Datasets.

preprint2011arXiv

Effects of gradient coupling on amplitude death in nonidentical oscillators

In this work, we investigate gradient coupling effect on amplitude death in an array of N cou- pled nonidentical oscillators with no-flux boundary conditions and periodic boundary conditions respectively. We find that the effects of gradient coupling on amplitude death in diffusive coupled nonidentical oscillators is quite different between those two boundaries conditions. With no-flux boundary conditions, there is a system size related critical gradient coupling $r_c$ within which the gradient coupling tends to monotonically enlarge the amplitude death domain in the parameter space. With the periodical boundary conditions, there is an optimal gradient coupling constant $r_o$ to realize largest AD domain. The gradient coupling first enlarges then decreases the amplitude death domain of diffusive coupled oscillators. The amplitude death domain of parameter space are analytically predicted for small number of gradient coupled oscillators.