|
| << previous |
Journal description
All volumes of the corresponding journal
All issues of the corresponding volume
All articles of the corresponding issues
Details for article 12 of 12 found articles
| Title: |
Mining maximal frequent patterns from univariate uncertain data |
| Author: |
Liu, Ying-Ho |
| Appeared in: |
Intelligent data analysis |
| Paging: | Volume 18 (2014) nr. 4 pages 653-676 |
| Year: |
2014-07-23 |
| Contents: |
In this paper, we propose mining maximal frequent patterns from univariate uncertain data. Univariate uncertain data refers to cases where each attribute in a transaction is associated with a quantitative interval and a probability density function that assigns a probability to each value in the interval. The number of frequent U2 patterns (i.e. frequent patterns of univariate uncertain data) is usually very large. To return a concise and informative mining result to users, we propose mining maximal frequent U2 patterns (MFU2Ps). A maximal frequent U2 pattern is a frequent U2 pattern without any frequent superset. The three proposed algorithms, MU2P-Miner, U2GenMax, and U2MAFIA, are different in terms of both the data formats used to store transactions and the structures used to store the MFU2Ps which are found during the mining process. The experiment results show that different algorithms excel when applied to different datasets and settings. We have applied the proposed algorithms to univariate uncertain data comprising measurements of the air quality and weather conditions in Taiwan; the derived MFU2Ps show that the air quality in Taiwan is usually good (unless a sand storm affects the island) and the weather is often wet. |
| Publisher: |
IOS Press |
| Source file: |
Elektronische Wetenschappelijke Tijdschriften |
Details for article 12 of 12 found articles
| << previous |