Thoughts on:
"Mining Inter-Transaction Associations with Templates",
Ling Feng, Hongjun Lu, Jeffery Xu Yu, Jiawei Han
Hong Kong Polytechnic University,
Hong Kong University of Science and Technology,
Australian National University, Simon Fraser University,
CIKM 1999
The template here is a template for a rule.
If each price/day/stock included more than the closing price, like the
price every 30 seconds, it would look more like our data. The interesting
thing in this paper is the idea of mining for rules across instances.
Their example:
If the prices of IBM and SUN go up, Microsoft's will most likely go up
the next day.
Here, each day is an instance and the movement of a stock's price is an
attribute. If one of their instances corresponds loosely to taking a
single time slice from our time series data, it might be possible
to adapt their method to our data. This used in conjunction with
similarity methods for finding support for rules found among all our
instances might be an interesting approach.
|