Development of Enhanced Neural Decision Tree Model and Application of Data Mining Techniques for Modeling of Petroleum Datasets

In this work, two data mining techniques are studied and applied to petroleum datasets. so as to extract useful information such as: correlations and interdependencies among the attributes. This helps to better understand important characteristics that can increase oil production. The two data mining techniques, Artificial Neural Networks and the Decision Tree algorithm, are combined to form the Neural Decision Tree (NDT) model. The NDT model enhances classification accuracy of the decision tree algorithm. The Neural Decision Tree model generates statistical results of classification and graphical representation of data in the form of a tree data structure, which supports data analysis in more detail, and useful knowledge can be extracted from it. The NDT model is useful for discovering interdependencies and for explaining correlations among the attributes of the data. The results generated from application of the Neural Decision Tree model can provide initial estimates of some parameters in the statistical modeling so that correlations among the parameters can be derived. The knowledge on correlations can support petroleum experts in making important decisions such as cost estimation and required resources for oil production.

Name thesis:degree_name

Master of Applied Science (MASc)

Level thesis:degree_level

Master's

Discipline thesis:degree_discipline

Engineering - Software Systems

Grantor dc:publisher

Faculty of Graduate Studies and Research, University of Regina

Year dc:date.issued

2014

Author dc:creator

• Omer, Sardar Usman

Advisor dc:contributor.advisor

• Chan, Christine

Committee members dc:contributor.committeemember

• Laforge, Paul
• Gelowitz, Craig

Language dc:language.iso

en

DOI

https://doi.org/10.82465/5023

OAI identifier oai:identifier

oai:uregina.scholaris.ca:10294/5776