Yue Zhao

Xueying Ding, Yue Zhao, Leman Akoglu

Outlier detection (OD) finds many applications with a rich literature of numerous techniques. Deep neural network based OD (DOD) has seen a recent surge of attention thanks to the many advances in deep learning. In this paper, we consider a critical-yet-understudied challenge with unsupervised DOD, that is, effective hyperparameter (HP) tuning/model selection. While several prior work report the sensitivity of OD models to HPs, it becomes ever so critical for the modern DOD models that exhibit a long list of HPs. We introduce HYPER for tuning DOD models, tackling two fundamental challenges: (1) validation without supervision (due to lack of labeled anomalies), and (2) efficient search of the HP/model space (due to exponential growth in the number of HPs). A key idea is to design and train a novel hypernetwork (HN) that maps HPs onto optimal weights of the main DOD model. In turn, HYPER capitalizes on a single HN that can dynamically generate weights for many DOD models (corresponding to varying HPs), which offers significant speed-up. In addition, it employs meta-learning on historical OD tasks with labels to train a proxy validation function, likewise trained with our proposed HN efficiently. Extensive experiments on 35 OD tasks show that HYPER achieves high performance against 8 baselines with significant efficiency gains.

Jaemin Yoo, Yue Zhao, Lingxiao Zhao et al.

Self-supervised learning (SSL) has proven effective in solving various problems by generating internal supervisory signals. Unsupervised anomaly detection, which faces the high cost of obtaining true labels, is an area that can greatly benefit from SSL. However, recent literature suggests that tuning the hyperparameters (HP) of data augmentation functions is crucial to the success of SSL-based anomaly detection (SSAD), yet a systematic method for doing so remains unknown. In this work, we propose DSV (Discordance and Separability Validation), an unsupervised validation loss to select high-performing detection models with effective augmentation HPs. DSV captures the alignment between an augmentation function and the anomaly-generating mechanism with surrogate losses, which approximate the discordance and separability of test data, respectively. As a result, the evaluation via DSV leads to selecting an effective SSAD model exhibiting better alignment, which results in high detection accuracy. We theoretically derive the degree of approximation conducted by the surrogate losses and empirically show that DSV outperforms a wide range of baselines on 21 real-world tasks.

Yuangang Li, Jiaqi Li, Zhuo Xiao et al.

Anomaly detection (AD) is an important machine learning task with applications in fraud detection, content moderation, and user behavior analysis. However, AD is relatively understudied in a natural language processing (NLP) context, limiting its effectiveness in detecting harmful content, phishing attempts, and spam reviews. We introduce NLP-ADBench, the most comprehensive NLP anomaly detection (NLP-AD) benchmark to date, which includes eight curated datasets and 19 state-of-the-art algorithms. These span 3 end-to-end methods and 16 two-step approaches that adapt classical, non-AD methods to language embeddings from BERT and OpenAI. Our empirical results show that no single model dominates across all datasets, indicating a need for automated model selection. Moreover, two-step methods with transformer-based embeddings consistently outperform specialized end-to-end approaches, with OpenAI embeddings outperforming those of BERT. We release NLP-ADBench at https://github.com/USC-FORTIS/NLP-ADBench, providing a unified framework for NLP-AD and supporting future investigations.

Haoyan Xu, Ruizhi Qian, Jiate Li et al.

Graph machine learning has advanced rapidly in tasks such as link prediction, anomaly detection, and node classification. As models scale up, pretrained graph models have become valuable intellectual assets because they encode extensive computation and domain expertise. Building on these advances, Graph Foundation Models (GFMs) mark a major step forward by jointly pretraining graph and text encoders on massive and diverse data. This unifies structural and semantic understanding, enables zero-shot inference, and supports applications such as fraud detection and biomedical analysis. However, the high pretraining cost and broad cross-domain knowledge in GFMs also make them attractive targets for model extraction attacks (MEAs). Prior work has focused only on small graph neural networks trained on a single graph, leaving the security implications for large-scale and multimodal GFMs largely unexplored. This paper presents the first systematic study of MEAs against GFMs. We formalize a black-box threat model and define six practical attack scenarios covering domain-level and graph-specific extraction goals, architectural mismatch, limited query budgets, partial node access, and training data discrepancies. To instantiate these attacks, we introduce a lightweight extraction method that trains an attacker encoder using supervised regression of graph embeddings. Even without contrastive pretraining data, this method learns an encoder that stays aligned with the victim text encoder and preserves its zero-shot inference ability on unseen graphs. Experiments on seven datasets show that the attacker can approximate the victim model using only a tiny fraction of its original training cost, with almost no loss in accuracy. These findings reveal that GFMs greatly expand the MEA surface and highlight the need for deployment-aware security defenses in large-scale graph learning systems.

Martin Q. Ma, Yue Zhao, Xiaorong Zhang et al.

Given an unsupervised outlier detection task, how should one select a detection algorithm as well as its hyperparameters (jointly called a model)? Unsupervised model selection is notoriously difficult, in the absence of hold-out validation data with ground-truth labels. Therefore, the problem is vastly understudied. In this work, we study the feasibility of employing internal model evaluation strategies for selecting a model for outlier detection. These so-called internal strategies solely rely on the input data (without labels) and the output (outlier scores) of the candidate models. We setup (and open-source) a large testbed with 39 detection tasks and 297 candidate models comprised of 8 detectors and various hyperparameter configurations. We evaluate 7 different strategies on their ability to discriminate between models w.r.t. detection performance, without using any labels. Our study reveals room for progress -- we find that none would be practically useful, as they select models only comparable to a state-of-the-art detector (with random configuration).

Yue Zhao, Zhi Qiao, Cao Xiao et al.

Despite the explosion of interest in healthcare AI research, the reproducibility and benchmarking of those research works are often limited due to the lack of standard benchmark datasets and diverse evaluation metrics. To address this reproducibility challenge, we develop PyHealth, an open-source Python toolbox for developing various predictive models on healthcare data. PyHealth consists of data preprocessing module, predictive modeling module, and evaluation module. The target users of PyHealth are both computer science researchers and healthcare data scientists. With PyHealth, they can conduct complex machine learning pipelines on healthcare datasets with fewer than ten lines of code. The data preprocessing module enables the transformation of complex healthcare datasets such as longitudinal electronic health records, medical images, continuous signals (e.g., electrocardiogram), and clinical notes into machine learning friendly formats. The predictive modeling module provides more than 30 machine learning models, including established ensemble trees and deep neural network-based approaches, via a unified but extendable API designed for both researchers and practitioners. The evaluation module provides various evaluation strategies (e.g., cross-validation and train-validation-test split) and predictive model metrics. With robustness and scalability in mind, best practices such as unit testing, continuous integration, code coverage, and interactive examples are introduced in the library's development. PyHealth can be installed through the Python Package Index (PyPI) or https://github.com/yzhao062/PyHealth .

Yue Zhao, Ryan A. Rossi, Leman Akoglu

Given an unsupervised outlier detection (OD) task on a new dataset, how can we automatically select a good outlier detection method and its hyperparameter(s) (collectively called a model)? Thus far, model selection for OD has been a "black art"; as any model evaluation is infeasible due to the lack of (i) hold-out data with labels, and (ii) a universal objective function. In this work, we develop the first principled data-driven approach to model selection for OD, called MetaOD, based on meta-learning. MetaOD capitalizes on the past performances of a large body of detection models on existing outlier detection benchmark datasets, and carries over this prior experience to automatically select an effective model to be employed on a new dataset without using any labels. To capture task similarity, we introduce specialized meta-features that quantify outlying characteristics of a dataset. Through comprehensive experiments, we show the effectiveness of MetaOD in selecting a detection model that significantly outperforms the most popular outlier detectors (e.g., LOF and iForest) as well as various state-of-the-art unsupervised meta-learners while being extremely fast. To foster reproducibility and further research on this new problem, we open-source our entire meta-learning system, benchmark environment, and testbed datasets.

Kwei-Herng Lai, Daochen Zha, Guanchu Wang et al.

We present TODS, an automated Time Series Outlier Detection System for research and industrial applications. TODS is a highly modular system that supports easy pipeline construction. The basic building block of TODS is primitive, which is an implementation of a function with hyperparameters. TODS currently supports 70 primitives, including data processing, time series processing, feature analysis, detection algorithms, and a reinforcement module. Users can freely construct a pipeline using these primitives and perform end- to-end outlier detection with the constructed pipeline. TODS provides a Graphical User Interface (GUI), where users can flexibly design a pipeline with drag-and-drop. Moreover, a data-driven searcher is provided to automatically discover the most suitable pipelines given a dataset. TODS is released under Apache 2.0 license at https://github.com/datamllab/tods.

Yue Zhao, Xiyang Hu, Cheng Cheng et al.

Outlier detection (OD) is a key machine learning (ML) task for identifying abnormal objects from general samples with numerous high-stake applications including fraud detection and intrusion detection. Due to the lack of ground truth labels, practitioners often have to build a large number of unsupervised, heterogeneous models (i.e., different algorithms with varying hyperparameters) for further combination and analysis, rather than relying on a single model. How to accelerate the training and scoring on new-coming samples by outlyingness (referred as prediction throughout the paper) with a large number of unsupervised, heterogeneous OD models? In this study, we propose a modular acceleration system, called SUOD, to address it. The proposed system focuses on three complementary acceleration aspects (data reduction for high-dimensional data, approximation for costly models, and taskload imbalance optimization for distributed environment), while maintaining performance accuracy. Extensive experiments on more than 20 benchmark datasets demonstrate SUOD's effectiveness in heterogeneous OD acceleration, along with a real-world deployment case on fraudulent claim analysis at IQVIA, a leading healthcare firm. We open-source SUOD for reproducibility and accessibility.

Yue Zhao, Zain Nasrullah, Zheng Li

PyOD is an open-source Python toolbox for performing scalable outlier detection on multivariate data. Uniquely, it provides access to a wide range of outlier detection algorithms, including established outlier ensembles and more recent neural network-based approaches, under a single, well-documented API designed for use by both practitioners and researchers. With robustness and scalability in mind, best practices such as unit testing, continuous integration, code coverage, maintainability checks, interactive examples and parallelization are emphasized as core components in the toolbox's development. PyOD is compatible with both Python 2 and 3 and can be installed through Python Package Index (PyPI) or https://github.com/yzhao062/pyod.

Lincan Li, Eren Erman Ozguven, Yue Zhao et al.

Accurate typhoon track forecasting is crucial for early system warning and disaster response. While Transformer-based models have demonstrated strong performance in modeling the temporal dynamics of dense trajectories of humans and vehicles in smart cities, they usually lack access to broader contextual knowledge that enhances the forecasting reliability of sparse meteorological trajectories, such as typhoon tracks. To address this challenge, we propose TyphoFormer, a novel framework that incorporates natural language descriptions as auxiliary prompts to improve typhoon trajectory forecasting. For each time step, we use Large Language Model (LLM) to generate concise textual descriptions based on the numerical attributes recorded in the North Atlantic hurricane database. The language descriptions capture high-level meteorological semantics and are embedded as auxiliary special tokens prepended to the numerical time series input. By integrating both textual and sequential information within a unified Transformer encoder, TyphoFormer enables the model to leverage contextual cues that are otherwise inaccessible through numerical features alone. Extensive experiments are conducted on HURDAT2 benchmark, results show that TyphoFormer consistently outperforms other state-of-the-art baseline methods, particularly under challenging scenarios involving nonlinear path shifts and limited historical observations.

Zheng Li, Yue Zhao, Xiyang Hu et al.

Outlier detection refers to the identification of data points that deviate from a general data distribution. Existing unsupervised approaches often suffer from high computational cost, complex hyperparameter tuning, and limited interpretability, especially when working with large, high-dimensional datasets. To address these issues, we present a simple yet effective algorithm called ECOD (Empirical-Cumulative-distribution-based Outlier Detection), which is inspired by the fact that outliers are often the "rare events" that appear in the tails of a distribution. In a nutshell, ECOD first estimates the underlying distribution of the input data in a nonparametric fashion by computing the empirical cumulative distribution per dimension of the data. ECOD then uses these empirical distributions to estimate tail probabilities per dimension for each data point. Finally, ECOD computes an outlier score of each data point by aggregating estimated tail probabilities across dimensions. Our contributions are as follows: (1) we propose a novel outlier detection method called ECOD, which is both parameter-free and easy to interpret; (2) we perform extensive experiments on 30 benchmark datasets, where we find that ECOD outperforms 11 state-of-the-art baselines in terms of accuracy, efficiency, and scalability; and (3) we release an easy-to-use and scalable (with distributed support) Python implementation for accessibility and reproducibility.

Kexin Huang, Tianfan Fu, Wenhao Gao et al.

Therapeutics machine learning is an emerging field with incredible opportunities for innovatiaon and impact. However, advancement in this field requires formulation of meaningful learning tasks and careful curation of datasets. Here, we introduce Therapeutics Data Commons (TDC), the first unifying platform to systematically access and evaluate machine learning across the entire range of therapeutics. To date, TDC includes 66 AI-ready datasets spread across 22 learning tasks and spanning the discovery and development of safe and effective medicines. TDC also provides an ecosystem of tools and community resources, including 33 data functions and types of meaningful data splits, 23 strategies for systematic model evaluation, 17 molecule generation oracles, and 29 public leaderboards. All resources are integrated and accessible via an open Python library. We carry out extensive experiments on selected datasets, demonstrating that even the strongest algorithms fall short of solving key therapeutics challenges, including real dataset distributional shifts, multi-scale modeling of heterogeneous data, and robust generalization to novel data points. We envision that TDC can facilitate algorithmic and scientific advances and considerably accelerate machine-learning model development, validation and transition into biomedical and clinical implementation. TDC is an open-science initiative available at https://tdcommons.ai.

Zheng Li, Yue Zhao, Jialin Fu

A synthetic dataset is a data object that is generated programmatically, and it may be valuable to creating a single dataset from multiple sources when direct collection is difficult or costly. Although it is a fundamental step for many data science tasks, an efficient and standard framework is absent. In this paper, we study a specific synthetic data generation task called downscaling, a procedure to infer high-resolution, harder-to-collect information (e.g., individual level records) from many low-resolution, easy-to-collect sources, and propose a multi-stage framework called SYNC (Synthetic Data Generation via Gaussian Copula). For given low-resolution datasets, the central idea of SYNC is to fit Gaussian copula models to each of the low-resolution datasets in order to correctly capture dependencies and marginal distributions, and then sample from the fitted models to obtain the desired high-resolution subsets. Predictive models are then used to merge sampled subsets into one, and finally, sampled datasets are scaled according to low-resolution marginal constraints. We make four key contributions in this work: 1) propose a novel framework for generating individual level data from aggregated data sources by combining state-of-the-art machine learning and statistical techniques, 2) perform simulation studies to validate SYNC's performance as a synthetic data generation algorithm, 3) demonstrate its value as a feature engineering tool, as well as an alternative to data collection in situations where gathering is difficult through two real-world datasets, 4) release an easy-to-use framework implementation for reproducibility and scalability at the production level that easily incorporates new data.

Zheng Li, Yue Zhao, Nicola Botta et al.

Outlier detection refers to the identification of rare items that are deviant from the general data distribution. Existing approaches suffer from high computational complexity, low predictive capability, and limited interpretability. As a remedy, we present a novel outlier detection algorithm called COPOD, which is inspired by copulas for modeling multivariate data distribution. COPOD first constructs an empirical copula, and then uses it to predict tail probabilities of each given data point to determine its level of "extremeness". Intuitively, we think of this as calculating an anomalous p-value. This makes COPOD both parameter-free, highly interpretable, and computationally efficient. In this work, we make three key contributions, 1) propose a novel, parameter-free outlier detection algorithm with both great performance and interpretability, 2) perform extensive experiments on 30 benchmark datasets to show that COPOD outperforms in most cases and is also one of the fastest algorithms, and 3) release an easy-to-use Python implementation for reproducibility.

Yue Zhao, Xueying Ding, Jianing Yang et al.

Outlier detection is a key field of machine learning for identifying abnormal data objects. Due to the high expense of acquiring ground truth, unsupervised models are often chosen in practice. To compensate for the unstable nature of unsupervised algorithms, practitioners from high-stakes fields like finance, health, and security, prefer to build a large number of models for further combination and analysis. However, this poses scalability challenges in high-dimensional large datasets. In this study, we propose a three-module acceleration framework called SUOD to expedite the training and prediction with a large number of unsupervised detection models. SUOD's Random Projection module can generate lower subspaces for high-dimensional datasets while reserving their distance relationship. Balanced Parallel Scheduling module can forecast the training and prediction cost of models with high confidence---so the task scheduler could assign nearly equal amount of taskload among workers for efficient parallelization. SUOD also comes with a Pseudo-supervised Approximation module, which can approximate fitted unsupervised models by lower time complexity supervised regressors for fast prediction on unseen data. It may be considered as an unsupervised model knowledge distillation process. Notably, all three modules are independent with great flexibility to "mix and match"; a combination of modules can be chosen based on use cases. Extensive experiments on more than 30 benchmark datasets have shown the efficacy of SUOD, and a comprehensive future development plan is also presented.

Yue Zhao, Maciej K. Hryniewicki

Selecting and combining the outlier scores of different base detectors used within outlier ensembles can be quite challenging in the absence of ground truth. In this paper, an unsupervised outlier detector combination framework called DCSO is proposed, demonstrated and assessed for the dynamic selection of most competent base detectors, with an emphasis on data locality. The proposed DCSO framework first defines the local region of a test instance by its k nearest neighbors and then identifies the top-performing base detectors within the local region. Experimental results on ten benchmark datasets demonstrate that DCSO provides consistent performance improvement over existing static combination approaches in mining outlying objects. To facilitate interpretability and reliability of the proposed method, DCSO is analyzed using both theoretical frameworks and visualization techniques, and presented alongside empirical parameter setting instructions that can be used to improve the overall performance.

Zain Nasrullah, Yue Zhao

Previous attempts at music artist classification use frame level audio features which summarize frequency content within short intervals of time. Comparatively, more recent music information retrieval tasks take advantage of temporal structure in audio spectrograms using deep convolutional and recurrent models. This paper revisits artist classification with this new framework and empirically explores the impacts of incorporating temporal structure in the feature representation. To this end, an established classification architecture, a Convolutional Recurrent Neural Network (CRNN), is applied to the artist20 music artist identification dataset under a comprehensive set of conditions. These include audio clip length, which is a novel contribution in this work, and previously identified considerations such as dataset split and feature level. Our results improve upon baseline works, verify the influence of the producer effect on classification performance and demonstrate the trade-offs between audio length and training set size. The best performing model achieves an average F1 score of 0.937 across three independent trials which is a substantial improvement over the corresponding baseline under similar conditions. Additionally, to showcase the effectiveness of the CRNN's feature extraction capabilities, we visualize audio samples at the model's bottleneck layer demonstrating that learned representations segment into clusters belonging to their respective artists.

Yue Zhao, Zain Nasrullah, Maciej K. Hryniewicki et al.

In unsupervised outlier ensembles, the absence of ground truth makes the combination of base outlier detectors a challenging task. Specifically, existing parallel outlier ensembles lack a reliable way of selecting competent base detectors, affecting accuracy and stability, during model combination. In this paper, we propose a framework---called Locally Selective Combination in Parallel Outlier Ensembles (LSCP)---which addresses the issue by defining a local region around a test instance using the consensus of its nearest neighbors in randomly selected feature subspaces. The top-performing base detectors in this local region are selected and combined as the model's final output. Four variants of the LSCP framework are compared with seven widely used parallel frameworks. Experimental results demonstrate that one of these variants, LSCP_AOM, consistently outperforms baselines on the majority of twenty real-world datasets.

Yue Zhao, Zhongtian Qiu, Yiqing Yang et al.

The emergence of smartwatches poses new challenges to information security. Although there are mature touch-based authentication methods for smartphones, the effectiveness of using these methods on smartwatches is still unclear. We conducted a user study (n=16) to evaluate how authentication methods (PIN and Pattern), UIs (Square and Circular), and display sizes (38mm and 42mm) affect authentication accuracy, speed, and security. Circular UIs are tailored to smartwatches with fewer UI elements. Results show that 1) PIN is more accurate and secure than Pattern; 2) Pattern is much faster than PIN; 3) Square UIs are more secure but less accurate than Circular UIs; 4) display size does not affect accuracy or speed, but security; 5) Square PIN is the most secure method of all. The study also reveals a security concern that participants' favorite method is not the best in any of the measures. We finally discuss implications for future touch-based smartwatch authentication design.