|Title:||Pattern mining algorithms : survey and improvement from association to taxonomy|
Hong Kong Polytechnic University -- Dissertations
|Department:||Department of Computing|
|Pages:||v, 94 leaves : ill. ; 30 cm.|
|Abstract:||This thesis gave us a complete conclusion on algorithms used in association mining problems. For a comparison, we introduced three different ways to pattern mining. (1) The Apriori algorithm, which is breadth-first-search algorithm, can be used in small problems where no special database is needed. However, we presented the solution for "maximal-clique" function and "subset" function, which can be used by other occasions. (2) The FP-growth algorithm, which mined the FP-tree without candidate generation, is the fastest and the most memory-efficient algorithm. We improved it with order code and prefix link generation. (3) The border-line algorithms, which is the simplest form of pattern mining, is concluded with two types of border-lines: the frequency line and the correlation line. A lot of graphs have been presented in this thesis - program constructs are also given in code form. The algorithms are constructed from real data and can run as the theoretical prediction. The taxonomy problem is thoroughly explored in the later sections. We first concluded two original ways, then discovered the quick method to mine taxonomically linked items without adding items or encode items. Besides these improvements and discussions, the DAG is researched with error checking and pseudo item value calculation. The database synthesis is also presented there where we solved the problem of pattern relation.|
Files in This Item:
|b25202248.pdf||For All Users (off-campus access for PolyU Staff & Students only)||1.34 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: