arXiv subjects

Che Lin

Publications and source records attributed to Che Lin.

17 recordsLinked to original sources

arXiv2026

Findings of the First Teaching Monster Challenge: A Benchmark of Pedagogical Content Knowledge in AI Agents

AI agents can now solve problems, answer like subject experts, and generate long-form multimodal content. However, whether they can adapt a lesson to fit a specified learner, which education calls Pedagogical Content Knowledge (PCK), has not been benchmarked. To measure it, we introduce the Teaching Monster Challenge, the first instructional video generation benchmark to treat the learner persona as an explicit evaluation criterion. Each system is given a topic and a learner persona and must generate a complete instructional video. Every video is screened by an LLM-judge, ranked by crowd pairwise voting, and finalized by an expert panel. The first edition shows that today's systems handle the content well but are far weaker at presenting it and adapting it to the learner. The same process exposes a limit of automatic judging. The LLM-judge separates a clear low-performing tail but ranks the strongest systems poorly. The strongest systems receive nearly identical scores from the judge, so its ranking of them does not match human preference. Progress therefore requires not only better teaching systems but also better automatic judges, and we release the benchmark, rubric, and human judgments as a testbed for both.

cs.AI↗

arXiv2026

Detecting Information Channels in Congressional Trading via Temporal Graph Learning

Congressional stock trading has raised concerns about potential information asymmetries and conflicts of interest in financial markets. We introduce a temporal graph network (TGN) framework to identify information channels through which members of Congress may possess advantageous knowledge when trading company stocks. We construct a multimodal dynamic graph integrating diverse publicly available datasets, including congressional stock transactions, lobbying relationships, campaign finance contributions, and geographical connections between legislators and corporations. Our approach formulates the detection problem as a dynamic edge classification task, where we identify trades that exhibit statistically significant outperformance relative to the S&P 500 across long time horizons. To handle the temporal nature of these relationships, we develop a two-step walk-forward validation architecture that respects information availability constraints and prevents look-ahead bias. We evaluate several labeling strategies based on risk-adjusted returns and demonstrate that the TGN successfully captures complex temporal dependencies between congressional-corporate interactions and subsequent trading performance.

cs.CE↗

arXiv2025

THeGAU: Type-Aware Heterogeneous Graph Autoencoder and Augmentation

Heterogeneous Graph Neural Networks (HGNNs) are effective for modeling Heterogeneous Information Networks (HINs), which encode complex multi-typed entities and relations. However, HGNNs often suffer from type information loss and structural noise, limiting their representational fidelity and generalization. We propose THeGAU, a model-agnostic framework that combines a type-aware graph autoencoder with guided graph augmentation to improve node classification. THeGAU reconstructs schema-valid edges as an auxiliary task to preserve node-type semantics and introduces a decoder-driven augmentation mechanism to selectively refine noisy structures. This joint design enhances robustness, accuracy, and efficiency while significantly reducing computational overhead. Extensive experiments on three benchmark HIN datasets (IMDB, ACM, and DBLP) demonstrate that THeGAU consistently outperforms existing HGNN methods, achieving state-of-the-art performance across multiple backbones.

cs.LG↗

arXiv2025

FinNuE: Exposing the Risks of Using BERTScore for Numerical Semantic Evaluation in Finance

BERTScore has become a widely adopted metric for evaluating semantic similarity between natural language sentences. However, we identify a critical limitation: BERTScore exhibits low sensitivity to numerical variation, a significant weakness in finance where numerical precision directly affects meaning (e.g., distinguishing a 2% gain from a 20% loss). We introduce FinNuE, a diagnostic dataset constructed with controlled numerical perturbations across earnings calls, regulatory filings, social media, and news articles. Using FinNuE, demonstrate that BERTScore fails to distinguish semantically critical numerical differences, often assigning high similarity scores to financially divergent text pairs. Our findings reveal fundamental limitations of embedding-based metrics for finance and motivate numerically-aware evaluation frameworks for financial NLP.

cs.CL↗

arXiv2025

MedFuse: Multiplicative Embedding Fusion For Irregular Clinical Time Series

Clinical time series derived from electronic health records (EHRs) are inherently irregular, with asynchronous sampling, missing values, and heterogeneous feature dynamics. While numerical laboratory measurements are highly informative, existing embedding strategies usually combine feature identity and value embeddings through additive operations, which constrains their ability to capture value-dependent feature interactions. We propose MedFuse, a framework for irregular clinical time series centered on the MuFuse (Multiplicative Embedding Fusion) module. MuFuse fuses value and feature embeddings through multiplicative modulation, preserving feature-specific information while modeling higher-order dependencies across features. Experiments on three real-world datasets covering both intensive and chronic care show that MedFuse consistently outperforms state-of-the-art baselines on key predictive tasks. Analysis of the learned representations further demonstrates that multiplicative fusion enhances expressiveness and supports cross-dataset pretraining. These results establish MedFuse as a generalizable approach for modeling irregular clinical time series.

cs.AI↗

arXiv2025

Learning from Limited Multi-Phase CT: Dual-Branch Prototype-Guided Framework for Early Recurrence Prediction in HCC

Early recurrence (ER) prediction after curative-intent resection remains a critical challenge in the clinical management of hepatocellular carcinoma (HCC). Although contrast-enhanced computed tomography (CT) with full multi-phase acquisition is recommended in clinical guidelines and routinely performed in many tertiary centers, complete phase coverage is not consistently available across all institutions. In practice, single-phase portal venous (PV) scans are often used alone, particularly in settings with limited imaging resources, variations in acquisition protocols, or patient-related factors such as contrast intolerance or motion artifacts. This variability results in a mismatch between idealized model assumptions and the practical constraints of real-world deployment, underscoring the need for methods that can effectively leverage limited multi-phase data. To address this challenge, we propose a Dual-Branch Prototype-guided (DuoProto) framework that enhances ER prediction from single-phase CT by leveraging limited multi-phase data during training. DuoProto employs a dual-branch architecture: the main branch processes single-phase images, while the auxiliary branch utilizes available multi-phase scans to guide representation learning via cross-domain prototype alignment. Structured prototype representations serve as class anchors to improve feature discrimination, and a ranking-based supervision mechanism incorporates clinically relevant recurrence risk factors. Extensive experiments demonstrate that DuoProto outperforms existing methods, particularly under class imbalance and missing-phase conditions. Ablation studies further validate the effectiveness of the dual-branch, prototype-guided design. Our framework aligns with current clinical application needs and provides a general solution for recurrence risk prediction in HCC, supporting more informed decision-making.

q-bio.QM↗

arXiv2025

Style4Rec: Enhancing Transformer-based E-commerce Recommendation Systems with Style and Shopping Cart Information

Understanding users' product preferences is essential to the efficacy of a recommendation system. Precision marketing leverages users' historical data to discern these preferences and recommends products that align with them. However, recent browsing and purchase records might better reflect current purchasing inclinations. Transformer-based recommendation systems have made strides in sequential recommendation tasks, but they often fall short in utilizing product image style information and shopping cart data effectively. In light of this, we propose Style4Rec, a transformer-based e-commerce recommendation system that harnesses style and shopping cart information to enhance existing transformer-based sequential product recommendation systems. Style4Rec represents a significant step forward in personalized e-commerce recommendations, outperforming benchmarks across various evaluation metrics. Style4Rec resulted in notable improvements: HR@5 increased from 0.681 to 0.735, NDCG@5 increased from 0.594 to 0.674, and MRR@5 increased from 0.559 to 0.654. We tested our model using an e-commerce dataset from our partnering company and found that it exceeded established transformer-based sequential recommendation benchmarks across various evaluation metrics. Thus, Style4Rec presents a significant step forward in personalized e-commerce recommendation systems.

cs.IR↗

arXiv2024

Scalable Numerical Embeddings for Multivariate Time Series: Enhancing Healthcare Data Representation Learning

Multivariate time series (MTS) data, when sampled irregularly and asynchronously, often present extensive missing values. Conventional methodologies for MTS analysis tend to rely on temporal embeddings based on timestamps that necessitate subsequent imputations, yet these imputed values frequently deviate substantially from their actual counterparts, thereby compromising prediction accuracy. Furthermore, these methods typically fail to provide robust initial embeddings for values infrequently observed or even absent within the training set, posing significant challenges to model generalizability. In response to these challenges, we propose SCAlable Numerical Embedding (SCANE), a novel framework that treats each feature value as an independent token, effectively bypassing the need for imputation. SCANE regularizes the traits of distinct feature embeddings and enhances representational learning through a scalable embedding mechanism. Coupling SCANE with the Transformer Encoder architecture, we develop the Scalable nUMerical eMbeddIng Transformer (SUMMIT), which is engineered to deliver precise predictive outputs for MTS characterized by prevalent missing entries. Our experimental validation, conducted across three disparate electronic health record (EHR) datasets marked by elevated missing value frequencies, confirms the superior performance of SUMMIT over contemporary state-of-the-art approaches addressing similar challenges. These results substantiate the efficacy of SCANE and SUMMIT, underscoring their potential applicability across a broad spectrum of MTS data analytical tasks.

cs.LG↗

arXiv2024

SynHING: Synthetic Heterogeneous Information Network Generation for Graph Learning and Explanation

Graph Neural Networks (GNNs) excel in delineating graph structures in diverse domains, including community analysis and recommendation systems. As the interpretation of GNNs becomes increasingly important, the demand for robust baselines and expansive graph datasets is accentuated, particularly in the context of Heterogeneous Information Networks (HIN). Addressing this, we introduce SynHING, a novel framework for Synthetic Heterogeneous Information Network Generation aimed at enhancing graph learning and explanation. SynHING systematically identifies major motifs in a target HIN and employs a bottom-up generation process with intra-cluster and inter-cluster merge modules. This process, supplemented by post-pruning techniques, ensures the synthetic HIN closely mirrors the original graph's structural and statistical properties. Crucially, SynHING provides ground-truth motifs for evaluating GNN explainer models, setting a new standard for explainable, synthetic HIN generation and contributing to the advancement of interpretable machine learning in complex networks.

cs.LG↗

arXiv2023

A GAN Approach for Node Embedding in Heterogeneous Graphs Using Subgraph Sampling

Graph neural networks (GNNs) face significant challenges with class imbalance, leading to biased inference results. To address this issue in heterogeneous graphs, we propose a novel framework that combines Graph Neural Network (GNN) and Generative Adversarial Network (GAN) to enhance classification for underrepresented node classes. The framework incorporates an advanced edge generation and selection module, enabling the simultaneous creation of synthetic nodes and edges through adversarial learning. Unlike previous methods, which predominantly focus on homogeneous graphs due to the difficulty of representing heterogeneous graph structures in matrix form, this approach is specifically designed for heterogeneous data. Existing solutions often rely on pre-trained models to incorporate synthetic nodes, which can lead to optimization inconsistencies and mismatches in data representation. Our framework avoids these pitfalls by generating data that aligns closely with the inherent graph topology and attributes, ensuring a more cohesive integration. Evaluations on multiple real-world datasets demonstrate the method's superiority over baseline models, particularly in tasks focused on identifying minority node classes, with notable improvements in performance metrics such as F-score and AUC-PRC score. These findings highlight the potential of this approach for addressing critical challenges in the field.

cs.LG↗

arXiv2015

Energy-Efficient Deterministic Adaptive Beamforming Algorithms for Distributed Sensor/Relay Networks

In this study, energy-efficient deterministic adaptive beamforming algorithms are proposed for distributed sensor/relay networks. Specifically, DBSA, D-QESA, D-QESA-E, and a hybrid algorithm, hybrid-QESA, that combines the benefits of both deterministic and random adaptive beamforming algorithms, are proposed. Rigorous convergence analyses are provided for all our proposed algorithms and convergence to the global optimal solution is shown for all our proposed algorithms. Through extensive numerical simulations, we demonstrate that superior performance is achieved by our proposed DBSA and D-QESA over random adaptive beamforming algorithms for static channels. Surprisingly, D-QESA is also more robust against random node removal than random adaptive beamforming algorithms. For time-varying channels, hybrid-QESA indeed achieves the best performance since it combines the benefits of both types of adaptive beamforming algorithms. In summary, our proposed deterministic algorithms demonstrate superior performance both in terms of convergence time and robustness against channel and network uncertainties.

cs.IT↗

arXiv2015

Deterministic Differential Search Algorithm for Distributed Sensor/Relay Networks

For distributed sensor/relay networks, high reliability and power efficiency are often required. However, several implementation issues arise in practice. One such problem is that all the distributed transmitters have limited power supply since the power source of the transmitters cannot be recharged continually. To resolve this, distributed beamforming has been proposed as a viable solution where all distributed transmitters seek to align in phase at the receiver end. However, it is difficult to implement such transmit beamforming in a distributed fashion in practice since perfect channel state information (CSI) need to be made available at all distributed transmitters, requiring tremendous overhead to feed back CSI from the receiver to all distributed transmitters. In this paper, we propose a novel algorithm that belongs to the category of deterministic phase adjustment algorithm: the Deterministic Differential Search Algorithm (DDSA), where the differences between the measured received signal strength (RSS) are utilized judiciously to help us predict the deterministic phase adjustment done at distributed transmitters. Numerical simulations demonstrate rapid convergence to a pre-determined threshold.

cs.IT↗

arXiv2011

Coordinated Beamforming for Multiuser MISO Interference Channel under Rate Outage Constraints

This paper studies the coordinated beamforming design problem for the multiple-input single-output (MISO) interference channel, assuming only channel distribution information (CDI) at the transmitters. Under a given requirement on the rate outage probability for receivers, we aim to maximize the system utility (e.g., the weighted sum rate, weighted geometric mean rate, and the weighed harmonic mean rate) subject to the rate outage constraints and individual power constraints. The outage constraints, however, lead to a complicated, nonconvex structure for the considered beamforming design problem and make the optimization problem difficult to handle. {Although} this nonconvex optimization problem can be solved in an exhaustive search manner, this brute-force approach is only feasible when the number of transmitter-receiver pairs is small. For a system with a large number of transmitter-receiver pairs, computationally efficient alternatives are necessary. The focus of this paper is hence on the design of such efficient approximation methods. In particular, by employing semidefinite relaxation (SDR) and first-order approximation techniques, we propose an efficient successive convex approximation (SCA) algorithm that provides high-quality approximate beamforming solutions via solving a sequence of convex approximation problems. The solution thus obtained is further shown to be a stationary point for the SDR of the original outage constrained beamforming design problem. {Furthermore}, we propose a distributed SCA algorithm where each transmitter optimizes its own beamformer using local CDI and information obtained from limited message exchange with the other transmitters. Our simulation results demonstrate that the proposed SCA algorithm and its distributed counterpart indeed converge, and near-optimal performance can be achieved for all the considered system utilities.

cs.IT↗

arXiv2010

A Bio-Inspired Robust Adaptive Random Search Algorithm for Distributed Beamforming

A bio-inspired robust adaptive random search algorithm (BioRARSA), designed for distributed beamforming for sensor and relay networks, is proposed in this work. It has been shown via a systematic framework that BioRARSA converges in probability and its convergence time scales linearly with the number of distributed transmitters. More importantly, extensive simulation results demonstrate that the proposed BioRARSA outperforms existing adaptive distributed beamforming schemes by as large as 29.8% on average. This increase in performance results from the fact that BioRARSA can adaptively adjust its sampling stepsize via the "swim" behavior inspired by the bacterial foraging mechanism. Hence, the convergence time of BioRARSA is insensitive to the initial sampling stepsize of the algorithm, which makes it robust against the dynamic nature of distributed wireless networks.

cs.IT↗

arXiv2010

A convex approximation approach to Weighted Sum Rate Maximization of Multiuser MISO Interference Channel under outage constraints

This paper considers weighted sum rate maximization of multiuser multiple-input single-output interference channel (MISO-IFC) under outage constraints. The outage-constrained weighted sum rate maximization problem is a nonconvex optimization problem and is difficult to solve. While it is possible to optimally deal with this problem in an exhaustive search manner by finding all the Pareto-optimal rate tuples in the (discretized) outage-constrained achievable rate region, this approach, however, suffers from a prohibitive computational complexity and is feasible only when the number of transmitter-receive pairs is small. In this paper, we propose a convex optimization based approximation method for efficiently handling the outage-constrained weighted sum rate maximization problem. The proposed approximation method consists of solving a sequence of convex optimization problems, and thus can be efficiently implemented by interior-point methods. Simulation results show that the proposed method can yield near-optimal solutions.

cs.IT↗

arXiv2010

A General Proof of Convergence for Adaptive Distributed Beamforming Schemes

This work focuses on the convergence analysis of adaptive distributed beamforming schemes that can be reformulated as local random search algorithms via a random search framework. Once reformulated as local random search algorithms, it is proved that under two sufficient conditions: a) the objective function of the algorithm is continuous and all its local maxima are global maxima, and b) the origin is an interior point within the range of the considered transformation of the random perturbation, the corresponding adaptive distributed beamforming schemes converge both in probability and in mean. This proof of convergence is general since it can be applied to analyze randomized adaptive distributed beamforming schemes with any type of objective functions and probability measures as long as both the sufficient conditions are satisfied. Further, this framework can be generalized to analyze an asynchronous scheme where distributed transmitters can only update their beamforming coefficients asynchronously. Simulation results are also provided to validate our analyses.

eess.SY↗

arXiv2007

To Code or Not to Code Across Time: Space-Time Coding with Feedback

Space-time codes leverage the availability of multiple antennas to enhance the reliability of communication over wireless channels. While space-time codes have initially been designed with a focus on open-loop systems, recent technological advances have enabled the possibility of low-rate feedback from the receiver to the transmitter. The focus of this paper is on the implications of this feedback in a single-user multi-antenna system with a general model for spatial correlation. We assume a limited feedback model, that is, a coherent receiver and statistics along with B bits of quantized channel information at the transmitter. We study space-time coding with a family of linear dispersion (LD) codes that meet an additional orthogonality constraint so as to ensure low-complexity decoding. Our results show that, when the number of bits of feedback (B) is small, a space-time coding scheme that is equivalent to beamforming and does not code across time is optimal in a weak sense in that it maximizes the average received SNR. As B increases, this weak optimality transitions to optimality in a strong sense which is characterized by the maximization of the average mutual information. Thus, from a system designer's perspective, our work suggests that beamforming may not only be attractive from a low-complexity viewpoint, but also from an information-theoretic viewpoint.

cs.IT↗