arXiv subjects

Chee Wei Tan

Publications and source records attributed to Chee Wei Tan.

At least 37 records · Page 2Linked to original sources

arXiv2025

DeepTrace: Learning to Optimize Contact Tracing in Epidemic Networks with Graph Neural Networks

Digital contact tracing aims to curb epidemics by identifying and mitigating public health emergencies through technology. Backward contact tracing, which tracks the sources of infection, proved crucial in places like Japan for identifying COVID-19 infections from superspreading events. This paper presents a novel perspective of digital contact tracing as online graph exploration and addresses the forward and backward contact tracing problem as a maximum-likelihood (ML) estimation problem using iterative epidemic network data sampling. The challenge lies in the combinatorial complexity and rapid spread of infections. We introduce DeepTrace, an algorithm based on a Graph Neural Network (GNN) that iteratively updates its estimations as new contact tracing data is collected, learning to optimize the maximum likelihood estimation by utilizing topological features to accelerate learning and improve convergence. The contact tracing process combines either BFS or DFS to expand the network and trace the infection source, ensuring comprehensive and efficient exploration. Additionally, the GNN model is fine-tuned through a two-phase approach: pre-training with synthetic networks to approximate likelihood probabilities and fine-tuning with high-quality data to refine the model. Using COVID-19 variant data, we illustrate that DeepTrace surpasses current methods in identifying superspreaders, providing a robust basis for a scalable digital contact tracing strategy.

cs.SI↗

arXiv2024

FedMoE-DA: Federated Mixture of Experts via Domain Aware Fine-grained Aggregation

Federated learning (FL) is a collaborative machine learning approach that enables multiple clients to train models without sharing their private data. With the rise of deep learning, large-scale models have garnered significant attention due to their exceptional performance. However, a key challenge in FL is the limitation imposed by clients with constrained computational and communication resources, which hampers the deployment of these large models. The Mixture of Experts (MoE) architecture addresses this challenge with its sparse activation property, which reduces computational workload and communication demands during inference and updates. Additionally, MoE facilitates better personalization by allowing each expert to specialize in different subsets of the data distribution. To alleviate the communication burdens between the server and clients, we propose FedMoE-DA, a new FL model training framework that leverages the MoE architecture and incorporates a novel domain-aware, fine-grained aggregation strategy to enhance the robustness, personalizability, and communication efficiency simultaneously. Specifically, the correlation between both intra-client expert models and inter-client data heterogeneity is exploited. Moreover, we utilize peer-to-peer (P2P) communication between clients for selective expert model synchronization, thus significantly reducing the server-client transmissions. Experiments demonstrate that our FedMoE-DA achieves excellent performance while reducing the communication pressure on the server.

cs.LG↗

arXiv2024

FedReMa: Improving Personalized Federated Learning via Leveraging the Most Relevant Clients

Federated Learning (FL) is a distributed machine learning paradigm that achieves a globally robust model through decentralized computation and periodic model synthesis, primarily focusing on the global model's accuracy over aggregated datasets of all participating clients. Personalized Federated Learning (PFL) instead tailors exclusive models for each client, aiming to enhance the accuracy of clients' individual models on specific local data distributions. Despite of their wide adoption, existing FL and PFL works have yet to comprehensively address the class-imbalance issue, one of the most critical challenges within the realm of data heterogeneity in PFL and FL research. In this paper, we propose FedReMa, an efficient PFL algorithm that can tackle class-imbalance by 1) utilizing an adaptive inter-client co-learning approach to identify and harness different clients' expertise on different data classes throughout various phases of the training process, and 2) employing distinct aggregation methods for clients' feature extractors and classifiers, with the choices informed by the different roles and implications of these model components. Specifically, driven by our experimental findings on inter-client similarity dynamics, we develop critical co-learning period (CCP), wherein we introduce a module named maximum difference segmentation (MDS) to assess and manage task relevance by analyzing the similarities between clients' logits of their classifiers. Outside the CCP, we employ an additional scheme for model aggregation that utilizes historical records of each client's most relevant peers to further enhance the personalization stability. We demonstrate the superiority of our FedReMa in extensive experiments.

cs.LG↗

arXiv2024

FedUHB: Accelerating Federated Unlearning via Polyak Heavy Ball Method

Federated learning facilitates collaborative machine learning, enabling multiple participants to collectively develop a shared model while preserving the privacy of individual data. The growing importance of the "right to be forgotten" calls for effective mechanisms to facilitate data removal upon request. In response, federated unlearning (FU) has been developed to efficiently eliminate the influence of specific data from the model. Current FU methods primarily rely on approximate unlearning strategies, which seek to balance data removal efficacy with computational and communication costs, but often fail to completely erase data influence. To address these limitations, we propose FedUHB, a novel exact unlearning approach that leverages the Polyak heavy ball optimization technique, a first-order method, to achieve rapid retraining. In addition, we introduce a dynamic stopping mechanism to optimize the termination of the unlearning process. Our extensive experiments show that FedUHB not only enhances unlearning efficiency but also preserves robust model performance after unlearning. Furthermore, the dynamic stopping mechanism effectively reduces the number of unlearning iterations, conserving both computational and communication resources. FedUHB can be proved as an effective and efficient solution for exact data removal in federated learning settings.

cs.LG↗

arXiv2024

Efficient Federated Unlearning with Adaptive Differential Privacy Preservation

Federated unlearning (FU) offers a promising solution to effectively address the need to erase the impact of specific clients' data on the global model in federated learning (FL), thereby granting individuals the ``Right to be Forgotten". The most straightforward approach to achieve unlearning is to train the model from scratch, excluding clients who request data removal, but it is resource-intensive. Current state-of-the-art FU methods extend traditional FL frameworks by leveraging stored historical updates, enabling more efficient unlearning than training from scratch. However, the use of stored updates introduces significant privacy risks. Adversaries with access to these updates can potentially reconstruct clients' local data, a well-known vulnerability in the privacy domain. While privacy-enhanced techniques exist, their applications to FU scenarios that balance unlearning efficiency with privacy protection remain underexplored. To address this gap, we propose FedADP, a method designed to achieve both efficiency and privacy preservation in FU. Our approach incorporates an adaptive differential privacy (DP) mechanism, carefully balancing privacy and unlearning performance through a novel budget allocation strategy tailored for FU. FedADP also employs a dual-layered selection process, focusing on global models with significant changes and client updates closely aligned with the global model, reducing storage and communication costs. Additionally, a novel calibration method is introduced to facilitate effective unlearning. Extensive experimental results demonstrate that FedADP effectively manages the trade-off between unlearning efficiency and privacy protection.

cs.CR↗

arXiv2024

Unraveling the Viral Spread of Misinformation: Maximum-Likelihood Estimation and Starlike Tree Approximation in Markovian Spreading Models

Identifying the source of epidemic-like spread in networks is crucial for removing internet viruses or finding the source of rumors in online social networks. The challenge lies in tracing the source from a snapshot observation of infected nodes. How do we accurately pinpoint the source? Utilizing snapshot data, we apply a probabilistic approach, focusing on the graph boundary and the observed time, to detect sources via an effective maximum likelihood algorithm. A novel starlike tree approximation extends applicability to general graphs, demonstrating versatility. Unlike previous works that rely heavily on structural properties alone, our method also incorporates temporal data for more precise source detection. We highlight the utility of the Gamma function for analyzing the ratio of the likelihood being the source between nodes asymptotically. Comprehensive evaluations confirm algorithmic effectiveness in diverse network scenarios, advancing source detection in large-scale network analysis and information dissemination strategies.

cs.SI↗

arXiv2024

A Systematic Survey and Critical Review on Evaluating Large Language Models: Challenges, Limitations, and Recommendations

Large Language Models (LLMs) have recently gained significant attention due to their remarkable capabilities in performing diverse tasks across various domains. However, a thorough evaluation of these models is crucial before deploying them in real-world applications to ensure they produce reliable performance. Despite the well-established importance of evaluating LLMs in the community, the complexity of the evaluation process has led to varied evaluation setups, causing inconsistencies in findings and interpretations. To address this, we systematically review the primary challenges and limitations causing these inconsistencies and unreliable evaluations in various steps of LLM evaluation. Based on our critical review, we present our perspectives and recommendations to ensure LLM evaluations are reproducible, reliable, and robust.

cs.CL↗

arXiv2024

FedDRL: A Trustworthy Federated Learning Model Fusion Method Based on Staged Reinforcement Learning

Traditional federated learning uses the number of samples to calculate the weights of each client model and uses this fixed weight value to fusion the global model. However, in practical scenarios, each client's device and data heterogeneity leads to differences in the quality of each client's model. Thus the contribution to the global model is not wholly determined by the sample size. In addition, if clients intentionally upload low-quality or malicious models, using these models for aggregation will lead to a severe decrease in global model accuracy. Traditional federated learning algorithms do not address these issues. To solve this probelm, we propose FedDRL, a model fusion approach using reinforcement learning based on a two staged approach. In the first stage, Our method could filter out malicious models and selects trusted client models to participate in the model fusion. In the second stage, the FedDRL algorithm adaptively adjusts the weights of the trusted client models and aggregates the optimal global model. We also define five model fusion scenarios and compare our method with two baseline algorithms in those scenarios. The experimental results show that our algorithm has higher reliability than other algorithms while maintaining accuracy.

cs.LG↗

arXiv2024

Towards Efficient and Certified Recovery from Poisoning Attacks in Federated Learning

Federated learning (FL) is vulnerable to poisoning attacks, where malicious clients manipulate their updates to affect the global model. Although various methods exist for detecting those clients in FL, identifying malicious clients requires sufficient model updates, and hence by the time malicious clients are detected, FL models have been already poisoned. Thus, a method is needed to recover an accurate global model after malicious clients are identified. Current recovery methods rely on (i) all historical information from participating FL clients and (ii) the initial model unaffected by the malicious clients, leading to a high demand for storage and computational resources. In this paper, we show that highly effective recovery can still be achieved based on (i) selective historical information rather than all historical information and (ii) a historical model that has not been significantly affected by malicious clients rather than the initial model. In this scenario, while maintaining comparable recovery performance, we can accelerate the recovery speed and decrease memory consumption. Following this concept, we introduce Crab, an efficient and certified recovery method, which relies on selective information storage and adaptive model rollback. Theoretically, we demonstrate that the difference between the global model recovered by Crab and the one recovered by train-from-scratch can be bounded under certain assumptions. Our empirical evaluation, conducted across three datasets over multiple machine learning models, and a variety of untargeted and targeted poisoning attacks reveals that Crab is both accurate and efficient, and consistently outperforms previous approaches in terms of both recovery speed and memory consumption.

cs.CR↗

arXiv2023

Large Language Model-Driven Classroom Flipping: Empowering Student-Centric Peer Questioning with Flipped Interaction

Reciprocal questioning is essential for effective teaching and learning, fostering active engagement and deeper understanding through collaborative interactions, especially in large classrooms. Can large language model (LLM), such as OpenAI's GPT (Generative Pre-trained Transformer) series, assist in this? This paper investigates a pedagogical approach of classroom flipping based on flipped interaction in LLMs. Flipped interaction involves using language models to prioritize generating questions instead of answers to prompts. We demonstrate how traditional classroom flipping techniques, including Peer Instruction and Just-in-Time Teaching (JiTT), can be enhanced through flipped interaction techniques, creating student-centric questions for hybrid teaching. In particular, we propose a workflow to integrate prompt engineering with clicker and JiTT quizzes by a poll-prompt-quiz routine and a quiz-prompt-discuss routine to empower students to self-regulate their learning capacity and enable teachers to swiftly personalize training pathways. We develop an LLM-driven chatbot software that digitizes various elements of classroom flipping and facilitates the assessment of students using these routines to deliver peer-generated questions. We have applied our LLM-driven chatbot software for teaching both undergraduate and graduate students from 2020 to 2022, effectively useful for bridging the gap between teachers and students in remote teaching during the COVID-19 pandemic years. In particular, LLM-driven classroom flipping can be particularly beneficial in large class settings to optimize teaching pace and enable engaging classroom experiences.

cs.CY↗

arXiv2023

Copilot for Xcode: Exploring AI-Assisted Programming by Prompting Cloud-based Large Language Models

This paper presents an AI-assisted programming tool called Copilot for Xcode for program composition and design to support human software developers. By seamlessly integrating cloud-based Large Language Models (LLM) with Apple's local development environment, Xcode, this tool enhances productivity and unleashes creativity for software development in Apple software ecosystem (e.g., iOS apps, macOS). Leveraging advanced natural language processing (NLP) techniques, Copilot for Xcode effectively processes source code tokens and patterns within code repositories, enabling features such as code generation, autocompletion, documentation, and error detection. Software developers can also query and make "small" decisions for program composition, some of which can be made simultaneously, and this is facilitated through prompt engineering in a chat interface of Copilot for Xcode. Finally, we present simple case studies as evidence of the effectiveness of utilizing NLP in Xcode to prompt popular LLM services like OpenAI ChatGPT for program composition and design.

cs.SE↗

arXiv2023

Contagion Source Detection in Epidemic and Infodemic Outbreaks: Mathematical Analysis and Network Algorithms

This monograph provides an overview of the mathematical theories and computational algorithm design for contagion source detection in large networks. By leveraging network centrality as a tool for statistical inference, we can accurately identify the source of contagions, trace their spread, and predict future trajectories. This approach provides fundamental insights into surveillance capability and asymptotic behavior of contagion spreading in networks. Mathematical theory and computational algorithms are vital to understanding contagion dynamics, improving surveillance capabilities, and developing effective strategies to prevent the spread of infectious diseases and misinformation.

cs.SI↗

arXiv2023

Natural Language Generation and Understanding of Big Code for AI-Assisted Programming: A Review

This paper provides a comprehensive review of the literature concerning the utilization of Natural Language Processing (NLP) techniques, with a particular focus on transformer-based large language models (LLMs) trained using Big Code, within the domain of AI-assisted programming tasks. LLMs, augmented with software naturalness, have played a crucial role in facilitating AI-assisted programming applications, including code generation, code completion, code translation, code refinement, code summarization, defect detection, and clone detection. Notable examples of such applications include the GitHub Copilot powered by OpenAI's Codex and DeepMind AlphaCode. This paper presents an overview of the major LLMs and their applications in downstream tasks related to AI-assisted programming. Furthermore, it explores the challenges and opportunities associated with incorporating NLP techniques with software naturalness in these applications, with a discussion on extending AI-assisted programming capabilities to Apple's Xcode for mobile software development. This paper also presents the challenges of and opportunities for incorporating NLP techniques with software naturalness, empowering developers with advanced coding assistance and streamlining the software development process.

cs.SE↗

arXiv2022

EuclidNet: Deep Visual Reasoning for Constructible Problems in Geometry

In this paper, we present a deep learning-based framework for solving geometric construction problems through visual reasoning, which is useful for automated geometry theorem proving. Constructible problems in geometry often ask for the sequence of straightedge-and-compass constructions to construct a given goal given some initial setup. Our EuclidNet framework leverages the neural network architecture Mask R-CNN to extract the visual features from the initial setup and goal configuration with extra points of intersection, and then generate possible construction steps as intermediary data models that are used as feedback in the training process for further refinement of the construction step sequence. This process is repeated recursively until either a solution is found, in which case we backtrack the path for a step-by-step construction guide, or the problem is identified as unsolvable. Our EuclidNet framework is validated on complex Japanese Sangaku geometry problems, demonstrating its capacity to leverage backtracking for deep visual reasoning of challenging problems.

cs.CV↗

arXiv2022

Epidemic Source Detection in Contact Tracing Networks: Epidemic Centrality in Graphs and Message-Passing Algorithms

We study the epidemic source detection problem in contact tracing networks modeled as a graph-constrained maximum likelihood estimation problem using the susceptible-infected model in epidemiology. Based on a snapshot observation of the infection subgraph, we first study finite degree regular graphs and regular graphs with cycles separately, thereby establishing a mathematical equivalence in maximal likelihood ratio between the case of finite acyclic graphs and that of cyclic graphs. In particular, we show that the optimal solution of the maximum likelihood estimator can be refined to distances on graphs based on a novel statistical distance centrality that captures the optimality of the nonconvex problem. An efficient contact tracing algorithm is then proposed to solve the general case of finite degree-regular graphs with multiple cycles. Our performance evaluation on a variety of graphs shows that our algorithms outperform the existing state-of-the-art heuristics using contact tracing data from the SARS-CoV 2003 and COVID-19 pandemics by correctly identifying the superspreaders on some of the largest superspreading infection clusters in Singapore and Taiwan.

cs.SI↗

arXiv2021

An Overview of Healthcare Data Analytics With Applications to the COVID-19 Pandemic

In the era of big data, standard analysis tools may be inadequate for making inference and there is a growing need for more efficient and innovative ways to collect, process, analyze and interpret the massive and complex data. We provide an overview of challenges in big data problems and describe how innovative analytical methods, machine learning tools and metaheuristics can tackle general healthcare problems with a focus on the current pandemic. In particular, we give applications of modern digital technology, statistical methods, data platforms and data integration systems to improve diagnosis and treatment of diseases in clinical research and novel epidemiologic tools to tackle infection source problems, such as finding Patient Zero in the spread of epidemics. We make the case that analyzing and interpreting big data is a very challenging task that requires a multi-disciplinary effort to continuously create more effective methodologies and powerful tools to transfer data information into knowledge that enables informed decision making.

cs.LG↗

arXiv2013

Joint Beamforming and Power Control in Coordinated Multicell: Max-Min Duality, Effective Network and Large System Transition

This paper studies joint beamforming and power control in a coordinated multicell downlink system that serves multiple users per cell to maximize the minimum weighted signal-to-interference-plus-noise ratio. The optimal solution and distributed algorithm with geometrically fast convergence rate are derived by employing the nonlinear Perron-Frobenius theory and the multicell network duality. The iterative algorithm, though operating in a distributed manner, still requires instantaneous power update within the coordinated cluster through the backhaul. The backhaul information exchange and message passing may become prohibitive with increasing number of transmit antennas and increasing number of users. In order to derive asymptotically optimal solution, random matrix theory is leveraged to design a distributed algorithm that only requires statistical information. The advantage of our approach is that there is no instantaneous power update through backhaul. Moreover, by using nonlinear Perron-Frobenius theory and random matrix theory, an effective primal network and an effective dual network are proposed to characterize and interpret the asymptotic solution.

cs.IT↗

arXiv2013

Rooting out the Rumor Culprit from Suspects

Suppose that a rumor originating from a single source among a set of suspects spreads in a network, how to root out this rumor source? With the a priori knowledge of suspect nodes and an observation of infected nodes, we construct a maximum a posteriori (MAP) estimator to identify the rumor source using the susceptible-infected (SI) model. The a priori suspect set and its associated connectivity bring about new ingredients to the problem, and thus we propose to use local rumor center, a generalized concept based on rumor centrality, to identify the source from suspects. For regular tree-type networks of node degree δ, we characterize Pc(n), the correct detection probability of the estimator upon observing n infected nodes, in both the finite and asymptotic regimes. First, when every infected node is a suspect, Pc(n) asymptotically grows from 0.25 to 0.307 with δ from 3 to infinity, a result first established in Shah and Zaman (2011, 2012) via a different approach; and it monotonically decreases with n and increases with δ. Second, when the suspects form a connected subgraph of the network, Pc(n) asymptotically significantly exceeds the a priori probability if δ>2, and reliable detection is achieved as δ becomes large; furthermore, it monotonically decreases with n and increases with δ. Third, when there are only two suspects, Pc(n) is asymptotically at least 0.75 if δ>2; and it increases with the distance between the two suspects. Fourth, when there are multiple suspects, among all possible connection patterns, that they form a connected subgraph of the network achieves the smallest detection probability. Our analysis leverages ideas from the Polya's urn model in probability theory and sheds insight into the behavior of the rumor spreading process not only in the asymptotic regime but also for the general finite-n regime.

cs.SI↗