UofA | Computing Science | Semester 2000-3 |
Advances in Frequent Pattern Mining (Independent Study)
Instructor: Osmar R. Zaïane
|
OBJECTIVE/DESCRIPTION:
This course will investigate approaches for mining frequent patterns
and compare different algorithms for association rule mining. The main
objective is to explore the different variations and optimizations of
the Apriori algorithm, study in depth some optimized algorithms, and
prospect the possibility of a one-pass-based approach for frequent
patterns mining that could be parallelized.
The course will consist of a series of discussions on recent relevant
research papers and findings, the testing of a variety of frequent
pattern mining approaches, the implementation of a new algorithm that
will be devised in the course of the study, and the preparation of a
final report with annotated bibliography.
TOPICS:
The course will cover the following topics:
- Frequent itemset counting.
- Apriori algorithm
- Multi-level association rule mining
- Quantitative association rule mining
- Parallel and distributed association rule
mining
- Frequent patterns with recurrent items
- Optimizations and knowledge
representation for frequent pattern
mining
GRADING:
Annotated bibliography (20%),
Discussions (20%),
Implementation and Testing (20%),
Final paper (40%).
TEXTBOOK and REFERENCES:
There is no textbook for this course. Research papers will be selected
from a variety of journals and conference proceedings, such as SIGKDD,
SIGMOD, ICDE, etc.
Distributed: July 25, 2000