Comparative Analysis of Taxi and SafeGraph Data for Understanding Human Mobility Patterns in New York City Neighborhoods
Taxi and SafeGraph data reveal distinct mobility patterns across New York City neighborhoods, highlighting the strengths and limitations of each dataset for understanding human mobility and transportation mode choices.
Attribute-Based Semantic Type Detection and Data Quality Assessment Using Attribute Labels for Enhanced Data Cleaning
Leveraging semantic information within attribute labels significantly enhances data quality assessment and streamlines the data cleaning process, leading to more efficient and effective data-driven decision-making.
利用形式概念分析為非標準數據設計數據深度函數
本文提出了一種利用形式概念分析為非標準數據定義數據深度函數的方法，並探討了其在識別數據中心和離群值方面的應用。
암호화폐 거래 네트워크에서 지역 및 글로벌 시간 모티프 마이닝의 통찰력과 주의 사항
이 논문에서는 암호화폐 거래 네트워크에서 시간적 모티프 분석을 사용하여 거래 패턴과 사용자 행동에 대한 통찰력을 얻는 방법을 제시합니다. 단순히 모티프 수를 세는 것만으로는 오해의 소지가 있는 결론을 도출할 수 있으며, 시간 경과에 따른 모티프 분포와 개별 노드에 대한 분석이 중요함을 강조합니다.
Efficient Mining of Weighted Sequential Patterns in Incremental Uncertain Databases
The core message of this work is to propose a novel framework for efficiently mining weighted sequential patterns in incremental uncertain databases. The framework introduces the concept of weighted expected support, along with several tightened upper bound measures and a hierarchical index structure to maintain patterns, enabling efficient mining of both unweighted and weighted uncertain sequential patterns.
Top-k Contrast Order-Preserving Pattern Mining: Efficient Discovery of Contrast Patterns for Time Series Classification
Discovering top-k contrast patterns for effective time series classification.
Identification and Uses of Deep Learning Backbones via Pattern Mining
Understanding and utilizing deep learning backbones for improved performance and explanation.
Top-k Contrast Order-Preserving Pattern Mining: Efficient Algorithm for Time Series Classification
Efficiently mine top-k contrast patterns for time series classification using the COPP-Miner algorithm.
Spectral Clustering of Categorical and Mixed-type Data with Extra Graph Nodes
The author proposes SpecMix, a spectral clustering algorithm that incorporates both numerical and categorical data by adding extra nodes to the graph. This approach leads to interpretable clustering results without the need for data preprocessing.
Efficient Top-k Contrast Order-Preserving Pattern Mining Algorithm
The author proposes the COPP-Miner algorithm for top-k contrast pattern mining to improve time series classification by discovering patterns with significant differences between classes efficiently.