• Publications
  • Influence
Efficient time series matching by wavelets
  • K. Chan, A. Fu
  • Computer Science
  • Proceedings 15th International Conference on Data…
  • 23 March 1999
This paper proposes to use Haar Wavelet Transform for time series indexing and shows that Haar transform can outperform DFT through experiments, and proposes a two-phase method for efficient n-nearest neighbor query in time series databases. Expand
HOT SAX: efficiently finding the most unusual time series subsequence
The utility of discords with objective experiments on domains as diverse as Space Shuttle telemetry monitoring, medicine, surveillance, and industry, and the effectiveness of the discord discovery algorithm with more than one million experiments, on 82 different datasets from diverse domains are demonstrated. Expand
Entropy-based subspace clustering for mining numerical data
This work considers a database with numerical attributes, in which each transaction is viewed as a multi-dimensional vector, and identifies new meaningful criteria of high density and correlation of dimensions for goodness of clustering in subspaces. Expand
Utility-based anonymization using local recoding
This paper proposes a simple framework to specify utility of attributes and develops two simple yet efficient heuristic local recoding methods for utility-based anonymization, which outperform the state-of-the-art multidimensional global recode methods in both discernability and query answering accuracy. Expand
Enhancing Effectiveness of Outlier Detections for Low Density Patterns
A connectivity-based outlier factor (COF) scheme is introduced that improves the effectiveness of an existing local outlier factors (LOF) scheme when a pattern itself has similar neighbourhood density as an outlier. Expand
Mining association rules with weighted items
Two new algorithms to mine the weighted association rules with weights, which make use of a metric called the k-support bound in the mining process, and show the efficiency of the algorithms for large databases. Expand
(α, k)-anonymity: an enhanced k-anonymity model for privacy preserving data publishing
It is proved that the optimal (α, k)-anonymity problem is NP-hard, and a local-recoding algorithm is proposed which is more scalable and result in less data distortion. Expand
A fast distributed algorithm for mining association rules
An interesting distributed association rule mining algorithm, FDM (fast distributed mining of association rules), which generates a small number of candidate sets and substantially reduces the number of messages to be passed at mining association rules is proposed. Expand
K-isomorphism: privacy preserving network publication against structural attacks
These investigations show that k-isomorphism, or anonymization by forming k pairwise isomorphic subgraphs, is both sufficient and necessary for the protection of privacy in social networks and the problem is shown to be NP-hard. Expand
Anonymizing transaction databases for publication
This paper proposes one way to address the problem of publishing "transaction data" for research purposes and proposes a satisfactory privacy notion and solution proposed for anonymizing transaction data. Expand