The Strength of Weak Learnability
Machine Learning
The nature of statistical learning theory
The nature of statistical learning theory
Fast Algorithms for Mining Association Rules in Large Databases
VLDB '94 Proceedings of the 20th International Conference on Very Large Data Bases
Consensus algorithms for the generation of all maximal bicliques
Discrete Applied Mathematics - The fourth international colloquium on graphs and optimisation (GO-IV)
Hi-index | 0.00 |
In this paper we describe a graph-theoretical approach for pattern discovery that is especially useful in epidemiological research when applied to case-control studies involving categorical features such as genotypes and exposures. Whereas existing approaches are limited to exploring relationships among two or three factors, or deal with thousands of genes but are unable to isolate interactions among individual genes, we focus on interactions among tens of genes. We present a pattern discovery algorithm that finds associations among multiple factors, such as genetic and environmental factors, and groups of individuals (cases and controls) in a clinical survey. To validate our approach and to demonstrate its effectiveness, we applied it to a selection of synthetic data sets that were devised to emulate the situations encountered in epidemiological studies involving common diseases with suspected associations involving multiple factors that could include inherited genotypes, somatic genotypes, demographic characteristics, or exposures. The results of this experiment show that it is possible to identify the effects of multiple factors in moderate-size surveys (involving hundreds of individuals) even when the number of factors is greater than three.