Authors

Han Yu

Type

Text

Type

Dissertation

Advisor

Zhu, Wei | Ahn, Hongshik | Wu, Song | Hong, Sangjin.

Date

2013-12-01

Keywords

Applied mathematics

Department

Department of Applied Mathematics and Statistics.

Language

en_US

Source

This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree.

Identifier

http://hdl.handle.net/11401/76460

Publisher

The Graduate School, Stony Brook University: Stony Brook, NY.

Format

application/pdf

Abstract

This dissertation presents applications of machine learning methods to clinical data sets and development of a decision support system. Our goal is to develop machine learning methods to predict potential healthcare problems before the onset of the actual diseases. Our research involves two examples with high-dimensional data. The independent variables are selected depending on the quality of prediction, and the models will be trained on the subspaces of the training data set. We also employed feature extraction technique to the original feature space such as PCA, FCA, data transformation, etc. This projection of the original feature space to lower the dimension is proven to be efficient in reducing the dimension of the data set. Recently, a non-probabilistic classifier called support vector machines (SVM) has been developed. The main idea of SVM is about mapping the input vectors into a high-dimensional feature space, and then a linear decision surface is constructed. Thus, the prediction will be based on the relative position of the data point with the decision surface. A tree-based ensemble method called Random Forest also received attention. Using a random selection of features to split each node yields error rates that make this method compare favorably to Adaboost. In this dissertation, we applied several machine learning methods and techniques to develop a reliable decision support system. | 97 pages

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.