Graph Neural Networks (GNNs) have received significant attention for demonstrating their capability to handle graph data. However, they are difficult to be deployed in resource-limited devices because of model sizes and scalability constraints imposed by the multi-hop data dependency. In addition, real-world graphs usually possess complex structural information and features. Therefore, to improve the applicability of GNNs and fully encode the complicated topological information, Knowledge Distillation on Graphs (KDG) has been introduced to build a smaller but effective model, leading to model compression and performance improvement. Recently, KDG has achieved considerable progress, with many studies proposed. In this survey, we systematically review these works. Specifically, we first introduce the challenges and bases of KDG, then categorize and summarize the existing work of KDG by answering the following three questions: (1) what to distillate, (2) who to whom, and (3) how to distillate. We offer in-depth comparisons and elucidate the strengths and weaknesses of each design. Finally, we share our thoughts on future research directions.
Rapid advancement in machine learning is increasing the demand for effective graph data analysis. However, real-world graph data often exhibits class imbalance, leading to poor performance of standard machine learning models on underrepresented classes. To address this, C lass- I mbalanced L earning on G raphs (CILG) has emerged as a promising solution that combines graph representation learning and class-imbalanced learning. This survey provides a comprehensive understanding of CILG’s current state-of-the-art, establishing the first systematic taxonomy of existing work and its connections to traditional imbalanced learning. We critically analyze recent advances and discuss key open problems. A continuously updated reading list of relevant articles and code implementations is available at https://github.com/yihongma/CILG-Papers .
Transferring the reasoning capability from stronger large language models (LLMs) to smaller ones has been quite appealing, as smaller LLMs are more flexible to deploy with less expense. Among the existing solutions, knowledge distillation stands out due to its outstanding efficiency and generalization. However, existing methods suffer from several drawbacks, including limited knowledge diversity and the lack of rich contextual information. To solve the problems and facilitate the learning of compact language models, we propose TinyLLM, a new knowledge distillation paradigm to learn a small student LLM from multiple large teacher LLMs. In particular, we encourage the student LLM to not only generate the correct answers but also understand the rationales behind these answers. Given that different LLMs possess diverse reasoning skills, we guide the student model to assimilate knowledge from various teacher LLMs. We further introduce an in-context example generator and a teacher-forcing Chain-of-Thought strategy to ensure that the rationales are accurate and grounded in contextually appropriate scenarios. Extensive experiments on six datasets across two reasoning tasks demonstrate the superiority of our method. Results show that TinyLLM can outperform large teacher LLMs significantly, despite a considerably smaller model size. The source code is available at: https://github.com/YikunHan42/TinyLLM.
Machine Unlearning (MU) algorithms have become increasingly critical due to the imperative adherence to data privacy regulations. The primary objective of MU is to erase the influence of specific data samples on a given model without the need to retrain it from scratch. Accordingly, existing methods focus on maximizing user privacy protection. However, there are different degrees of privacy regulations for each real-world web-based application. Exploring the full spectrum of trade-offs between privacy, model utility, and runtime efficiency is critical for practical unlearning scenarios. Furthermore, designing the MU algorithm with simple control of the aforementioned trade-off is desirable but challenging due to the inherent complex interaction. To address the challenges, we present Controllable Machine Unlearning (ConMU), a novel framework designed to facilitate the calibration of MU. The ConMU framework contains three integral modules: an important data selection module that reconciles the runtime efficiency and model generalization, a progressive Gaussian mechanism module that balances privacy and model generalization, and an unlearning proxy that controls the trade-offs between privacy and runtime efficiency. Comprehensive experiments on various benchmark datasets have demonstrated the robust adaptability of our control mechanism and its superiority over established unlearning methods. ConMU explores the full spectrum of the Privacy-Utility-Efficiency trade-off and allows practitioners to account for different real-world regulations. Source code available at: https://github.com/guangyaodou/ConMU
31 December 2024 in Expert Systems with Applications