Database Reference
In-Depth Information
more comprehensible models than weighted literals, such as decision trees
or rules, and literals over intervals can also be used with these models.
A first attempt to learn rules of literals is described in [Rodrıguez
et al.
(2000)], where we obtained less accurate but more comprehensible clas-
sifiers. Nevertheless, the use of these literals with other learning models
requires more research.
Finally, a short remark about the suitability of the method for time
series classification. Since the only properties of the series that are consid-
ered for their classification are the attributes tested by the interval based lit-
erals, the method will be adequate as long as the series may be discriminated
according to what happens in intervals. Hence, although the experimental
results are rather good, they cannot be generalized for arbitrary data sets.
Nevertheless, this learning framework can be used with other kinds of lit-
erals, more suited for the problem at hand. For instance [Rodrıguez and
Alonso (2000)] uses literals based on distances between series.
References
1. Alcock, R.J. and Manolopoulos, Y. (1999). Time-Series Similarity Queries
Employing a Feature-Based Approach. In
Proceedings of the 7th Hellenic
Conference on Informatics, Ioannina, Greece
.
2. Alonso Gonzalez, C.J. and Rodr ıguez Diez, J.J. (1999). A Graphical
Rule Language for Continuous Dynamic Systems. In
International Confer-
ence on Computational Intelligence for Modelling, Control and Automation
(CIMCA'99), Vienna, Austria
.
3. Berndt, D. and Clifford, J. (1996). Finding Patterns in Time Series: A
Dynamic Programming Approach. In Fayyad, U., Piatetsky-Shapiro, G.,
Smyth, P., and Uthurusamy, R., eds,
Advances in Knowledge Discovery and
Data Mining
, pp. 229-248. AAAI Press/MIT Press.
4. Hettich,
S.
and
Bay,
S.D.
(1999).
The
UCI
KDD
archive.
http://kdd.ics.uci.edu/.
5. Kadous, M.W. (1999). Learning Comprehensible Descriptions of Multivariate
Time Series. In Bratko, I. and Dzeroski, S., eds,
Proceedings of the 16th Inter-
national Conference of Machine Learning (ICML-99)
. Morgan Kaufmann.
6. Kubat, M., Koprinska, I., and Pfurtscheller, G. (1998). Learning to Classify
Biomedical Signals. In Michalski, R., Bratko, I., and Kubat, M., eds, Machine
Learning and Data Mining, pp. 409-428. John Wiley & Sons.
7. Rodr ıguez, J.J. and Alonso, C.J. (2000). Applying Boosting to Similarity
Literals for Time Series Classification. In
Proceedings of the 1st International
Workshop on Multiple Classifier Systems (MCS 2000)
. Springer.
8. Rodr ıguez, J.J., Alonso, C.J., and Bostrom, H. (2000). Learning First Order
Logic Time Series Classifiers: Rules and Boosting. In
Proceedings of the 4th
European Conference on Principles of Data Mining and Knowledge Discovery
(PKDD 2000)
. Springer.
