arXiv subjects

Yunlong Ma

Publications and source records attributed to Yunlong Ma.

10 recordsLinked to original sources

arXiv2025

R-Tuning: Wavelet-Decomposed Replay and Semantic Alignment for Continual Adaptation of Pretrained Time-Series Models

Pre-trained models have demonstrated exceptional generalization capabilities in time-series forecasting; however, adapting them to evolving data distributions remains a significant challenge. A key hurdle lies in accessing the original training data, as fine-tuning solely on new data often leads to catastrophic forgetting. To address this issue, we propose Replay Tuning (R-Tuning), a novel framework designed for the continual adaptation of pre-trained time-series models. R-Tuning constructs a unified latent space that captures both prior and current task knowledge through a frequency-aware replay strategy. Specifically, it augments model-generated samples via wavelet-based decomposition across multiple frequency bands, generating trend-preserving and fusion-enhanced variants to improve representation diversity and replay efficiency. To further reduce reliance on synthetic samples, R-Tuning introduces a latent consistency constraint that aligns new representations with the prior task space. This constraint guides joint optimization within a compact and semantically coherent latent space, ensuring robust knowledge retention and adaptation. Extensive experimental results demonstrate the superiority of R-Tuning, which reduces MAE and MSE by up to 46.9% and 46.8%, respectively, on new tasks, while preserving prior knowledge with gains of up to 5.7% and 6.0% on old tasks. Notably, under few-shot settings, R-Tuning outperforms all state-of-the-art baselines even when synthetic proxy samples account for only 5% of the new task dataset.

cs.LG↗

arXiv2025

A Multi-Memory Segment System for Generating High-Quality Long-Term Memory Content in Agents

In the current field of agent memory, extensive explorations have been conducted in the area of memory retrieval, yet few studies have focused on exploring the memory content. Most research simply stores summarized versions of historical dialogues, as exemplified by methods like A-MEM and MemoryBank. However, when humans form long-term memories, the process involves multi-dimensional and multi-component generation, rather than merely creating simple summaries. The low-quality memory content generated by existing methods can adversely affect recall performance and response quality. In order to better construct high-quality long-term memory content, we have designed a multi-memory segment system (MMS) inspired by cognitive psychology theory. The system processes short-term memory into multiple long-term memory segments, and constructs retrieval memory units and contextual memory units based on these segments, with a one-to-one correspondence between the two. During the retrieval phase, MMS will match the most relevant retrieval memory units based on the user's query. Then, the corresponding contextual memory units is obtained as the context for the response stage to enhance knowledge, thereby effectively utilizing historical data. We conducted experiments on the LoCoMo dataset and further performed ablation experiments, experiments on the robustness regarding the number of input memories, and overhead experiments, which demonstrated the effectiveness and practical value of our method.

cs.AI↗

arXiv2025

Agent That Debugs: Dynamic State-Guided Vulnerability Repair

In recent years, more vulnerabilities have been discovered every day, while manual vulnerability repair requires specialized knowledge and is time-consuming. As a result, many detected or even published vulnerabilities remain unpatched, thereby increasing the exposure of software systems to attacks. Recent advancements in agents based on Large Language Models have demonstrated their increasing capabilities in code understanding and generation, which can be promising to achieve automated vulnerability repair. However, the effectiveness of agents based on static information retrieval is still not sufficient for patch generation. To address the challenge, we propose a program repair agent called VulDebugger that fully utilizes both static and dynamic context, and it debugs programs in a manner akin to humans. The agent inspects the actual state of the program via the debugger and infers expected states via constraints that need to be satisfied. By continuously comparing the actual state with the expected state, it deeply understands the root causes of the vulnerabilities and ultimately accomplishes repairs. We experimentally evaluated VulDebugger on 50 real-life projects. With 60.00% successfully fixed, VulDebugger significantly outperforms state-of-the-art approaches for vulnerability repair.

cs.SE↗

arXiv2025

NodeNAS: Node-Specific Graph Neural Architecture Search for Out-of-Distribution Generalization

Graph neural architecture search (GraphNAS) has demonstrated advantages in mitigating performance degradation of graph neural networks (GNNs) due to distribution shifts. Recent approaches introduce weight sharing across tailored architectures, generating unique GNN architectures for each graph end-to-end. However, existing GraphNAS methods do not account for distribution patterns across different graphs and heavily rely on extensive training data. With sparse or single training graphs, these methods struggle to discover optimal mappings between graphs and architectures, failing to generalize to out-of-distribution (OOD) data. In this paper, we propose node-specific graph neural architecture search(NodeNAS), which aims to tailor distinct aggregation methods for different nodes through disentangling node topology and graph distribution with limited datasets. We further propose adaptive aggregation attention based Multi-dim NodeNAS method(MNNAS), which learns an node-specific architecture customizer with good generalizability. Specifically, we extend the vertical depth of the search space, supporting simultaneous node-specific architecture customization across multiple dimensions. Moreover, we model the power-law distribution of node degrees under varying assortativity, encoding structure invariant information to guide architecture customization across each dimension. Extensive experiments across supervised and unsupervised tasks demonstrate that MNNAS surpasses state-of-the-art algorithms and achieves excellent OOD generalization.

cs.LG↗

arXiv2024

Apollo-Forecast: Overcoming Aliasing and Inference Speed Challenges in Language Models for Time Series Forecasting

Encoding time series into tokens and using language models for processing has been shown to substantially augment the models' ability to generalize to unseen tasks. However, existing language models for time series forecasting encounter several obstacles, including aliasing distortion and prolonged inference times, primarily due to the limitations of quantization processes and the computational demands of large models. This paper introduces Apollo-Forecast, a novel framework that tackles these challenges with two key innovations: the Anti-Aliasing Quantization Module (AAQM) and the Race Decoding (RD) technique. AAQM adeptly encodes sequences into tokens while mitigating high-frequency noise in the original signals, thus enhancing both signal fidelity and overall quantization efficiency. RD employs a draft model to enable parallel processing and results integration, which markedly accelerates the inference speed for long-term predictions, particularly in large-scale models. Extensive experiments on various real-world datasets show that Apollo-Forecast outperforms state-of-the-art methods by 35.41\% and 18.99\% in WQL and MASE metrics, respectively, in zero-shot scenarios. Furthermore, our method achieves a 1.9X-2.7X acceleration in inference speed over baseline methods.

cs.LG↗

arXiv2024

Synchro-Transient-Extracting Transform for the Analysis of Signals with Both Harmonic and Impulsive Components

Time-frequency analysis (TFA) techniques play an important role in the field of machine fault diagnosis attributing to their superiority in dealing with nonstationary signals. Synchroextracting transform (SET) and transient-extracting transform (TET) are two newly emerging techniques that can produce energy concentrated representation for nonstationary signals. However, SET and TET are only suitable for processing harmonic signals and impulsive signals, respectively. This poses a challenge for each of these two techniques when a signal contains both harmonic and impulsive components. In this paper, we propose a new TFA technique to solve this problem. The technique first divides time-frequency (TF) coefficients into two types according to the chirp rate of the signal, and then TF ridges are respectively extracted to prevent the limitation for the independent use of SET and TET. This allows the technique to generate an energy-concentrated TF representation for both harmonics and impulses in the signal. Furthermore, we theoretically demonstrate that the proposed technique retains the signal reconstruction capability. The effectiveness of the proposed technique is verified using numerical and real-world signals.

eess.SP↗

arXiv2022

FastATDC: Fast Anomalous Trajectory Detection and Classification

Automated detection of anomalous trajectories is an important problem with considerable applications in intelligent transportation systems. Many existing studies have focused on distinguishing anomalous trajectories from normal trajectories, ignoring the large differences between anomalous trajectories. A recent study has made great progress in identifying abnormal trajectory patterns and proposed a two-stage algorithm for anomalous trajectory detection and classification (ATDC). This algorithm has excellent performance but suffers from a few limitations, such as high time complexity and poor interpretation. Here, we present a careful theoretical and empirical analysis of the ATDC algorithm, showing that the calculation of anomaly scores in both stages can be simplified, and that the second stage of the algorithm is much more important than the first stage. Hence, we develop a FastATDC algorithm that introduces a random sampling strategy in both stages. Experimental results show that FastATDC is 10 to 20 times faster than ATDC on real datasets. Moreover, FastATDC outperforms the baseline algorithms and is comparable to the ATDC algorithm.

cs.LG↗

arXiv2021

Automated Allocation of Detention Rooms Based on Inverse Graph Partitioning

Room allocation is a challenging task in detention centers since lots of related people need to be held separately with limited rooms. It is extremely difficult and risky to allocate rooms manually, especially for organized crime groups with close connections. To tackle this problem, we develop an intelligent room allocation system for detention centers to provide optimized room allocation schemes automatically. We first formalize the detention room allocation problem as inverse graph partitioning, which can measure the quality of room allocation schemes. Then, we propose two heuristic algorithms to achieve the global optimization and local optimization of detention room allocation. Experiment results on real-world datasets show that the proposed algorithms significantly outperform manual allocation and suggest that the system is of great practical application value.

cs.SI↗

arXiv2021

Towards Fast Evaluation of Unsupervised Link Prediction by Random Sampling Unobserved Links

Link prediction has aroused extensive attention since it can both discover hidden connections and predict future links in the networks. Many unsupervised link prediction algorithms have been proposed to find these links in a variety of networks. However, there is an evaluation conundrum in unsupervised link prediction. Unobserved links heavily outnumber observed links in large networks, so it is unrealistic to quantify the existence likelihood of all unobserved links to evaluate these algorithms. In this paper, we propose a new evaluation paradigm that is sampling unobserved links to address this problem. First, we demonstrate that the proposed paradigm is feasible in theory. Then, we perform extensive evaluation experiments in real-world networks of different contexts and sizes. The results show that the performance of similarity-based link prediction algorithms is highly stable even at a low sampling ratio in large networks, and the evaluation time degradation caused by sampling is striking. Our findings have broad implications for link prediction.

cs.SI↗

arXiv2018

SdcNet: A Computation-Efficient CNN for Object Recognition

Extracting features from a huge amount of data for object recognition is a challenging task. Convolution neural network can be used to meet the challenge, but it often requires a large number of computation resources. In this paper, a computation-efficient convolutional module, named SdcBlock, is proposed and based on it, the convolution network SdcNet is introduced for object recognition tasks. In the proposed module, optimized successive depthwise convolutions supported by appropriate data management is applied in order to generate vectors containing high density and more varieties of feature information. The hyperparameters can be easily adjusted to suit varieties of tasks under different computation restrictions without significantly jeopardizing the performance. The experiments have shown that SdcNet achieved an error rate of 5.60% in CIFAR-10 with only 55M Flops and also reduced further the error rate to 5.24% using a moderate volume of 103M Flops. The expected computation efficiency of the SdcNet has been confirmed.

eess.IV↗