C4.5: programs for machine learning
C4.5: programs for machine learning
Machine Learning
Feature selection in unsupervised learning via evolutionary search
Proceedings of the sixth ACM SIGKDD international conference on Knowledge discovery and data mining
Text Categorization with Suport Vector Machines: Learning with Many Relevant Features
ECML '98 Proceedings of the 10th European Conference on Machine Learning
Constrained K-means Clustering with Background Knowledge
ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
Temporal Data Mining Using Hidden Markov-Local Polynomial Models
PAKDD '01 Proceedings of the 5th Pacific-Asia Conference on Knowledge Discovery and Data Mining
Feature selection in data mining
Data mining
Toward Integrating Feature Selection Algorithms for Classification and Clustering
IEEE Transactions on Knowledge and Data Engineering
Kernel Based Algorithms for Mining Huge Data Sets: Supervised, Semi-supervised, and Unsupervised Learning (Studies in Computational Intelligence)
An application of element oriented analysis based credit scoring
ICDM'10 Proceedings of the 10th industrial conference on Advances in data mining: applications and theoretical aspects
A stratified model for short-term prediction of time series
PRICAI'10 Proceedings of the 11th Pacific Rim international conference on Trends in artificial intelligence
Hi-index | 0.00 |
Mining multidimensional data has two major concerns. One is how to select the most salient attributes and another one is how to guarantee the precision of mining results. This paper introduces a novel approach to mine multidimensional data through Element Oriented Analysis (EOA). In our approach, each observational data is considered to be comprised by two essential elements, the structure elements and the numerical elements. EOA firstly targets Structural Element Pattern (SEP) that is an aggregation of the structural elements. The successful SEP will be referenced by a Numerical Element Pattern (NEP) that is composed of the numerical elements. Given the results from both SEP and NEP, a global discriminant will be created for the efficient evaluation by the consequent data. In this paper, publicly available Turkish bank records are analyzed in an experiment that demonstrates the practical utility of our approach.