Yue Wu

Yue Wu


Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 July 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
JCJunda Chen
Junda Chen
AMAlessandro Muscoloni
Alessandro Muscoloni
IAIlyes Abdelhamid
Ilyes Abdelhamid
YWYue Wu
Yue Wu
Carlo Vittorio CannistraciCarlo Vittorio Cannistraci
Carlo Vittorio Cannistraci
The AUC-ROC is a widespread measure to assess the performance of a binary classification model across various discrimination thresholds. Yet, it is not apt for unbalanced class problems and is misleading for early retrieval. For instance, link prediction in complex networks is an unbalanced early retrieval problem, whose goal is to offer a ranking of the nonobserved network links, prioritizing a small cohort of positive links (for which we are sure the labels are reliable) on top of a list largely populated by unlabeled links (for which the labels are uncertain). Differently from binary classification, here the evaluation focuses on how the predictor prioritizes the positive class, because the negative class does not exist and is replaced by an unlabeled class. For decades, scholars have been investigating the necessity to correct and to generalize the AUC-ROC to work for both unbalanced classification and early retrieval evaluation, this is requested in many applied domains of computational science. Here we propose the area under the magnified ROC (AUC-mROC), a new measure that adjusts the standard AUC-ROC to work for unbalanced early retrieval problems such as link prediction. Finally, we introduce the area under the generalized ROC (AUC-gROC), which unifies both AUC-mROC and AUC-ROC to provide a universal and reliable evaluation measure that can work for both classification and early retrieval problems, regardless of class proportion.This discovery represents a critical achievement, culminating in success after decades of contentious research. Its solution promises to significantly impact various domains within applied computational science.

29 January 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
YWYue Wu
Yue Wu
Carlo Vittorio CannistraciCarlo Vittorio Cannistraci
Carlo Vittorio Cannistraci
Imbalanced data pose a challenge to the evaluation of trained machine-learning models due to evaluation bias, leading to inappropriate models for real-world applications. Data augmentation serves as a possible solution, but it introduces uncertainty due to the augmentation strategy and the data quality. Comparing measures which rely on diverse statistical principles to assess classification performance in imbalanced data is beneficial in this uncertainty situations. Here, we approach this challenge proposing the accuracy score (AC-score) which combines sensitivity and specificity in an unbiased measure with two relevant properties. First, AC-score is symmetric with respect to both positive and negative classes, this implies that AC-score adapts to cases where both classes have equal importance. Second, AC-score penalizes more than other existing measures the models whose prediction favors one class over the other. We show that AC-score is more conservative than geometric mean, AUC-ROC and balanced accuracy, and we comment the cases in which this offers an advantage offering empirical evidence on artificial and real data classifications.

29 January 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
YWYue Wu
Yue Wu
Carlo Vittorio CannistraciCarlo Vittorio Cannistraci
Carlo Vittorio Cannistraci
Scalability is an important aspect of algorithm design. Typical methods for accessing scalability in terms of running time include theoretical time complexity by algorithm analysis, and model-fitting time complexity derived from simulation results. However, theoretical time complexity often fails to account for real-world conditions in algorithm implementation such as influence from compilers, and lacks a simulation-based examination method. Alternatively, model-fitting time complexity is prone to learn a model with incorrect scaling orders. Here, we propose the empirical time complexity (ETC), which is a data-driven model-free and parameter-free method to account for the factors influencing an algorithm; for instance, the crosstalk between algorithm realization, compilation and hardware implementation. This can be used to guide code optimization and to gain algorithmic maximum efficiency for specific hardware, programming language or compiler. The values of ETC in function of an input variable form a curve that offers a visual representation of how an algorithm scale with input size. When the ETC curves of different versions of an algorithm are reported together with the theoretical time complexity curve, their comparison allows to select what versions are closer to the theoretical complexity. To showcase the utility of ETC in real scenarios, we investigated two sets of algorithms (graph shortest path and matrix multiplication) and we offer evidence on how ETC is crucial for diagnostic and optimized design of algorithms as close as possible to their theoretical limit.

21 June 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
AAAldo Acevedo
Aldo Acevedo
YWYue Wu
Yue Wu
FTFabio Lorenzo Traversa
Fabio Lorenzo Traversa
Carlo Vittorio CannistraciCarlo Vittorio Cannistraci
Carlo Vittorio Cannistraci
Complexity science studies physical phenomena that cannot be explained by the mere analysis of the single units of a system but requires to account for their interactions. A feature of complexity in connected systems is the emergence of mesoscale patterns in a geometric space, such as groupings in bird flocks. These patterns are formed by groups of points that tend to separate from each other, creating mesoscale structures. When multidimensional data or complex networks are embedded in a geometric space, some mesoscale patterns can appear respectively as clusters or communities, and their geometric separability is a feature according to which the performance of an algorithm for network embedding can be evaluated. Here, we introduce a framework for the definition and measure of the geometric separability (linear and nonlinear) of mesoscale patterns by solving the travelling salesman problem (TSP), and we offer experimental evidence on embedding and visualization of multidimensional data or complex networks, which are generated artificially or are derived from real complex systems. For the first time in literature the TSP’s solution is used to define a criterion of nonlinear separability of points in a geometric space, hence redefining the separability problem in terms of the travelling salesman problem is an innovation which impacts both computer science and complexity theory.

2 April 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
YWYue Wu
Yue Wu
JZJialin Zhao
Jialin Zhao
Carlo Vittorio CannistraciCarlo Vittorio Cannistraci
Carlo Vittorio Cannistraci
With the advent of the neural network era, traditional machine learning methods have increasingly been overshadowed. Nevertheless, continuing research on the role of geometry for learning in data science is crucial to envision and understand new principles behind the design of efficient machine learning. Linear classifiers are favored in certain tasks due to their reduced susceptibility to overfitting and their ability to provide interpretable decision boundaries. However, achieving both scalability and high predictive performance in linear classification remains a persistent challenge. Here, we propose a theoretical framework named geometric discriminant analysis (GDA). GDA includes the family of linear classifiers that can be expressed as a function of a centroid discriminant basis (CDB0) - the connection line between two centroids - adjusted by geometric corrections under different constraints. We demonstrate that linear discriminant analysis (LDA) is a subcase of the GDA theoretical framework, and we show its convergence to CDB0 under certain conditions. Then, based on the GDA framework, we propose an efficient linear classifier named centroid discriminant analysis (CDA) which is defined as a special case of GDA under a 2D plane geometric constraint. CDA training is initialized starting from CDB0 and involves the iterative calculation of new adjusted centroid discriminant lines whose optimal rotations on the associated 2D planes are searched via Bayesian optimization. CDA has good scalability (quadratic time complexity) which is lower than LDA and support vector machine (SVM) (cubic complexity). Results on 27 real datasets across classification tasks of standard images, medical images and chemical properties, offer empirical evidence that CDA outperforms other linear methods such as LDA, SVM and logistic regression (LR) in terms of scalability, performance and stability. Furthermore, we show that linear CDA can be generalized to nonlinear CDA via kernel method, demonstrating improvements on the linear version with tests on three challenging datasets of images and chemical data. GDA's general validity as a new theoretical framework may inspire the design of new classifiers under the definition of different geometric constraints, paving the way towards a deeper understanding of the role of geometry in learning from data.