Mining MOUCLAS Patterns and Jumping MOUCLAS Patterns to Construct Classifiers - Data Mining: Theory, Methodology, Techniques, and Applications - page 128

oil/gas formation since. In this sense, a selected set of such MPs or JMPs can be a
useful guide to petroleum engineers to identify possible drilling targets and their
depth and thickness at the stage of exploration and exploitation.
MPs and JMPs aim at deriving an explicit or implicit heuristic relationship between
measured values (well logging data) and properties to be predicted (oil/gas formation or
not). The MOUCLAS based method is ideally suitable to establish such implicit
relationships through proper training. The notable advantage of MOUCLAS based
algorithms over more traditional processing techniques such as model based well logging
analysis is that a physical model to describe the relationship between the well logging
data and the property of interest is not needed; nor is an very precise understanding of the
physical phenomena of the well logging data. From this point of view, MOUCLAS based
algorithms provides a complementary and useful technical approach towards the
interpretation of petroleum data and benefits petroleum discovery.
6 Conclusions
Two novel classification patterns, the MOUCLAS Pattern ( MP ) and the Jumping
MOUCLAS Pattern ( JMP ) for quantitative data in high dimensional databases, are
investigated in this paper. We also propose the algorithm for the discovery of the
interesting MPs and JMPs and construct two new classifiers called De-MP and J-MP .
As a hybrid of classification and clustering and association rules mining, our approach
may have several advantages which are (1) it has a solid mathematical foundation and
compact mathematical description of classifiers, (2) it does not require discretization,
as opposed to other, otherwise quite similar methods such as ARCS are strongly
related to, (3) it is robust when handling noisy or incomplete data in high dimensional
data space, regardless of the database size, due to its grid-based characteristic. An
illustration of application of MPs and JMPs is presented for the cost effective and
intelligent well logging data analysis for reservoir characterization. In the future
research, we attempt to carry out experiments on petroleum datasets to establish a
relationship between different well logs, seismic attributes, laboratory measurements
and other reservoir properties to evaluate performance of the MOUCLAS algorithms
proposed in this paper.
Acknowledgement
This work was partially supported by the Australia-China Special Fund for Scientific
and Technological Cooperation under grant CH030086.
References
1. Fayyad, U. M., Piatetsky-Shapiro, G., & Smyth, P. From data mining to knowledge
discovery: An overview. Advances in knowledge discovery and data mining. AAAI/MIT
Press. (1996) 1-34
2. Han, J., & M. Kamber. Data mining: concepts and techniques. Morgan Kaufmann
Publishers. (2000)