A5101641280- Imbalanced Data Classification Techniques
- Recommender Systems and Techniques
- Advanced Algebra and Geometry
- Algebraic Geometry and Number Theory
- Circular RNAs in diseases
- Artificial Intelligence in Healthcare
- Anomaly Detection Techniques and Applications
- Machine Learning and Data Classification
- MicroRNA in disease regulation
- Electricity Theft Detection Techniques
- Privacy-Preserving Technologies in Data
- Network Security and Intrusion Detection
- Ferroptosis and cancer prognosis
- Ethics and Social Impacts of AI
- Advanced Bandit Algorithms Research
- Financial Distress and Bankruptcy Prediction
- Geometry and complex manifolds
- Rough Sets and Fuzzy Logic
- Biological Activity of Diterpenoids and Biflavonoids
- Mobile Agent-Based Network Management
- Advanced Data Storage Technologies
- UAV Applications and Optimization
- Advanced Banach Space Theory
- Advanced Causal Inference Techniques
- Domain Adaptation and Few-Shot Learning
Jinzhou Medical University
2024-2025
University of Illinois Urbana-Champaign
2022-2024
Laboratoire Jean-Alexandre Dieudonné
2024
China Pharmaceutical University
2024
Institut de recherche mathématique de Rennes
2022-2023
Université de Rennes
2023
Centre National de la Recherche Scientifique
2023
Zhejiang Financial College
2021
Jilin Medical University
2020-2021
Jilin University
2017-2021
Finding node correspondence across networks, namely multi-network alignment, is an essential prerequisite for joint learning on multiple networks. Despite great success in aligning networks pairs, the literature alignment sparse due to exponentially growing solution space and lack of high-order discrepancy measures. To fill this gap, we propose a hierarchical multi-marginal optimal transport framework named HOT alignment. handle large space, are decomposed into smaller aligned clusters via...
Many real-world applications reveal difficulties in learning classifiers from imbalanced data. The rising big data era has been witnessing more classification tasks with large-scale but extremely imbalance and low-quality datasets. Most of existing methods suffer poor performance or low computation efficiency under such a scenario. To tackle this problem, we conduct deep investigations into the nature class imbalance, which reveals that not only disproportion between classes, also other...
Imbalanced learning (IL), i.e., unbiased models from class-imbalanced data, is a challenging problem. Typical IL methods including resampling and reweighting were designed based on some heuristic assumptions. They often suffer unstable performance, poor applicability, high computational cost in complex tasks where their assumptions do not hold. In this paper, we introduce novel ensemble framework named MESA. It adaptively resamples the training set iterations to get multiple classifiers...
Ensuring equitable impact of machine learning models across different societal groups is utmost importance for real-world applications. Prior research in fairness has predominantly focused on adjusting model outputs through pre-processing, in-processing, or post-processing techniques. These techniques focus correcting bias either the data model. However, we argue that and should be addressed conjunction. To achieve this, propose an algorithm called GroupDebias to reduce unfairness a...
Tn this paper we present a new design of an autorecharge drone system consist drones auto-landing program and recharging ground stations, working with battery swapping charging structure. Recently the application expanded fast, for aerial photography or other entertainment use are no longer only thing micro can do. Companies like Amazon starting to consider in their day-to-day business, also Chinese ministry electric power already have some special built quadcopter-drone do wire inspection...
Abstract We study the classification problem for polarized varieties with high nef value. give a complete list of isomorphism classes normal This generalizes classical work on smooth case by Fujita, Beltrametti and Sommese. As consequence we obtain that slc singularities value are birationally equivalent to projective bundles over nodal curves.
Ecto-5-nucleotidase (CD73) is overexpressed in a variety of cancers and associated with the immunosuppressive tumor microenvironment, making it an attractive target for cancer immunotherapy. Herein, we designed synthesized series novel (pyridazine-3-yl)pyrimidine-2,4(1
Class-imbalance is a common problem in machine learning practice. Typical Imbalanced Learning (IL) methods balance the data via intuitive class-wise resampling or reweighting. However, previous studies suggest that beyond class-imbalance, intrinsic difficulty factors like overlapping, noise, and small disjuncts also play critical roles. To handle them, many solutions have been proposed (e.g., noise removal, borderline sampling, hard example mining) but are still confined to specific factor...
Malicious cybersecurity activities have become increasingly worrisome for individuals and companies alike. While machine learning methods like Graph Neural Networks (GNNs) proven successful on the malware detection task, their output is often difficult to understand. Explainable are needed automatically identify malicious programs present results analysts in a way that human interpretable. In this survey, we outline number of GNN explainability compare performance real-world dataset....
imbalanced-ensemble, abbreviated as imbens, is an open-source Python toolbox for leveraging the power of ensemble learning to address class imbalance problem. It provides standard implementations popular imbalanced (EIL) methods with extended features and utility functions. These include resampling-based, e.g., under/over-sampling, reweighting-based, cost-sensitive learning. Beyond implementation, we empower EIL algorithms new functionalities like customizable resampling scheduler verbose...
Uplift modeling aims to model the incremental impact of a treatment on an individual outcome, which has attracted great interests researchers and practitioners from different communities. Existing uplift methods rely either data collected randomized controlled trials (RCTs) or observational is more realistic. However, we notice that data, it often case only small number subjects receive treatment, but finally infer much large group subjects. Such highly imbalanced common in various fields...
Unsupervised Anomaly Detection (UAD) is a key data mining problem owing to its wide real-world applications. Due the complete absence of supervision signals, UAD methods rely on implicit assumptions about anomalous patterns (e.g., scattered/sparsely/densely clustered) detect anomalies. However, are complex and vary significantly across different domains. No single assumption can describe such complexity be valid in all scenarios. This also confirmed by recent research that shows no method...
Data collected in the real world often encapsulates historical discrimination against disadvantaged groups and individuals. Existing fair machine learning (FairML) research has predominantly focused on mitigating discriminative bias model prediction, with far less effort dedicated towards exploring how to trace biases present data, despite its importance for transparency interpretability of FairML. To fill this gap, we investigate a novel problem: discovering samples that reflect...
Watermarking of large language models (LLMs) generation embeds an imperceptible statistical pattern within texts, making it algorithmically detectable. is a promising method for addressing potential harm and biases from LLMs, as enables traceability, accountability, detection manipulated content, helping to mitigate unintended consequences. However, open-source models, watermarking faces two major challenges: (i) incompatibility with fine-tuned (ii) vulnerability fine-tuning attacks. In this...
Many payment platforms hold large-scale marketing campaigns, which allocate incentives to encourage users pay through their applications. To maximize the return on investment, incentive allocations are commonly solved in a two-stage procedure. After training response estimation model estimate users' mobile probabilities (MPP), linear programming process is applied obtain optimal allocation. However, large amount of biased data set, generated by previous allocation policy, causes estimation....
Many payment platforms hold large-scale marketing campaigns, which allocate incentives to encourage users pay through their applications. To maximize the return on investment, incentive allocations are commonly solved in a two-stage procedure. After training response estimation model estimate users' mobile probabilities (MPP), linear programming process is applied obtain optimal allocation. However, large amount of biased data set, generated by previous allocation policy, causes estimation....
Unsupervised Anomaly Detection (UAD) is a key data mining problem owing to its wide real-world applications. Due the complete absence of supervision signals, UAD methods rely on implicit assumptions about anomalous patterns (e.g., scattered/sparsely/densely clustered) detect anomalies. However, are complex and vary significantly across different domains. No single assumption can describe such complexity be valid in all scenarios. This also confirmed by recent research that shows no method...