How Developers Discuss Generative AI: A Longitudinal Study of the Visual Studio Code Community
研究通过分析Visual Studio Code社区的讨论，使用关键词检索和语义过滤结合的方法，探讨了开发者如何讨论生成式AI工具及其在软件开发中的实际应用问题。
0 citationsRead paper
Selected work
Representative Papers
Robust Bichromatic Classification in 3D Using Planes and Slices
This study addresses the problem of robust binary classification of red and blue point sets in three-dimensional space, with a focus on resilience against various types of outliers. Leveraging the duality between points and planes in 3D, the work extends existing two-dimensional robust bichromatic classification techniques to the three-dimensional setting and formulates a classifier centered on linear constraints. The authors develop several efficient geometric algorithms tailored to distinct outlier models, including spatial outliers and label noise. Experimental results demonstrate that the proposed approach consistently and accurately separates red and blue point sets across diverse contamination scenarios, significantly enhancing the practicality and adaptability of robust classification in three dimensions.
0 citationsRead paper
Covering Planar Lattices with Interior-Disjoint Unit Disks
This study investigates the feasibility of covering periodic point sets in the plane—specifically triangular, square, and honeycomb lattices—with non-overlapping unit disks, determining the ranges of inter-point distances that permit complete coverage. By constructing single- and multi-family periodic covering patterns centered at lattice vertices, face centers, and off-lattice points, and combining symmetry analysis with computer-assisted verification, the work systematically characterizes the conditions for full coverage. The main contributions include identifying several new intervals of coverable spacing for the triangular lattice, establishing the first effective covering scheme for honeycomb point sets, and correcting an overlap error in prior results for the square lattice—thereby recovering known coverable intervals and discovering additional ones.
0 citationsRead paper
An Improved Linear Extractable Sketch Data Structure for Flow Count Statistics
This work addresses a key limitation in traditional linear sketching methods for streaming data statistics, which rely on the strong assumption that each hash bucket contains only a single key—thereby constraining space efficiency. To overcome this, the authors propose a novel approach that stores randomized linear combinations of multiple keys within each bucket and reconstructs key-value pairs during recovery by solving a sparse linear system. This design effectively relaxes the single-key-per-bucket constraint, achieving substantially improved space efficiency with only a modest increase in computational overhead. Experimental results demonstrate that the proposed method significantly reduces memory consumption while markedly enhancing space utilization for streaming data statistics.
0 citationsRead paper
LGTD: Local-Global Trend Decomposition for Season-Length-Free Time Series Analysis
This work proposes an adaptive time series decomposition framework that eliminates the need to predefine or estimate seasonal periods, addressing the limitations of traditional methods in handling non-stationary, drifting, or multi-scale seasonal patterns. The approach decomposes a sequence into three components: a global trend, an adaptive local linear trend that implicitly captures seasonality, and a residual. An AutoTrend module dynamically partitions the local trend in an error-driven manner, while global smoothing ensures coherent long-term structure; seasonality emerges automatically as a recurring pattern in the local trends. Operating in linear time, the method demonstrates robust performance across synthetic datasets with fixed, transitioning, and varying seasonal periods, achieving high-quality, low-intervention decomposition even in scenarios where conventional techniques fail.
0 citationsRead paper
Recent publications
Latest Papers
How Developers Discuss Generative AI: A Longitudinal Study of the Visual Studio Code Community
研究通过分析Visual Studio Code社区的讨论，使用关键词检索和语义过滤结合的方法，探讨了开发者如何讨论生成式AI工具及其在软件开发中的实际应用问题。
0 citationsRead paper
Robust Bichromatic Classification in 3D Using Planes and Slices
This study addresses the problem of robust binary classification of red and blue point sets in three-dimensional space, with a focus on resilience against various types of outliers. Leveraging the duality between points and planes in 3D, the work extends existing two-dimensional robust bichromatic classification techniques to the three-dimensional setting and formulates a classifier centered on linear constraints. The authors develop several efficient geometric algorithms tailored to distinct outlier models, including spatial outliers and label noise. Experimental results demonstrate that the proposed approach consistently and accurately separates red and blue point sets across diverse contamination scenarios, significantly enhancing the practicality and adaptability of robust classification in three dimensions.
0 citationsRead paper
Covering Planar Lattices with Interior-Disjoint Unit Disks
This study investigates the feasibility of covering periodic point sets in the plane—specifically triangular, square, and honeycomb lattices—with non-overlapping unit disks, determining the ranges of inter-point distances that permit complete coverage. By constructing single- and multi-family periodic covering patterns centered at lattice vertices, face centers, and off-lattice points, and combining symmetry analysis with computer-assisted verification, the work systematically characterizes the conditions for full coverage. The main contributions include identifying several new intervals of coverable spacing for the triangular lattice, establishing the first effective covering scheme for honeycomb point sets, and correcting an overlap error in prior results for the square lattice—thereby recovering known coverable intervals and discovering additional ones.
0 citationsRead paper
An Improved Linear Extractable Sketch Data Structure for Flow Count Statistics
This work addresses a key limitation in traditional linear sketching methods for streaming data statistics, which rely on the strong assumption that each hash bucket contains only a single key—thereby constraining space efficiency. To overcome this, the authors propose a novel approach that stores randomized linear combinations of multiple keys within each bucket and reconstructs key-value pairs during recovery by solving a sparse linear system. This design effectively relaxes the single-key-per-bucket constraint, achieving substantially improved space efficiency with only a modest increase in computational overhead. Experimental results demonstrate that the proposed method significantly reduces memory consumption while markedly enhancing space utilization for streaming data statistics.
0 citationsRead paper
LGTD: Local-Global Trend Decomposition for Season-Length-Free Time Series Analysis
This work proposes an adaptive time series decomposition framework that eliminates the need to predefine or estimate seasonal periods, addressing the limitations of traditional methods in handling non-stationary, drifting, or multi-scale seasonal patterns. The approach decomposes a sequence into three components: a global trend, an adaptive local linear trend that implicitly captures seasonality, and a residual. An AutoTrend module dynamically partitions the local trend in an error-driven manner, while global smoothing ensures coherent long-term structure; seasonality emerges automatically as a recurring pattern in the local trends. Operating in linear time, the method demonstrates robust performance across synthetic datasets with fixed, transitioning, and varying seasonal periods, achieving high-quality, low-intervention decomposition even in scenarios where conventional techniques fail.
0 citationsRead paper