Journal Name:
- European Journal of Pure and Applied Mathematics
Keywords (Original Language):
Author Name | University of Author | Faculty of Author |
---|---|---|
Bookmark/Search this post with
Abstract (Original Language):
In this paper we develop a novel tree based radial basis function neural networks (RBF-NNs)
model incorporating logistic regression. We aim to improve the classification performance of logistic
regression method by pre-processing the input data in RBF-NN frame. Although the scope of our proposed
method is binary classification in this paper, it is easy to generalize it for multi-class classification
problems. Furthermore, our model is very convenient to adapt for n < p classification problem that is
very popular yet difficult topic in statistics. We show the generalization and classification performance
of our model using simulated data. We have also applied our model on a real life data set gathered
from hydraulic fracturing in Oil & Gas wells. The results show the high classification performance of
our model that is superior to logistic regression. We have coded our model on R software. Logistic
Regression applications were carried out using IBM SPSS Version 20.
FULL TEXT (PDF):
- 4