Semantically enabled neural network modeling of major depressive disorder

Xiao, Yuezhang
Thumbnail Image
Other Contributors
Embrechts, Mark J.
Chan, Wai Kin (Victor)
Mitchell, John E.
Issue Date
Industrial Systems engineering
Terms of Use
Attribution-NonCommercial-NoDerivs 3.0 United States
This electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
Full Citation
We then describe our user case: we combine this portal with our second order differential equation and neural network model for depressive data analysis. Both the design methods and learning algorithm are discussed in this thesis. Here we focus on how different antidepressant treatments, including non-pharmacological treatments and psychotherapy, affect clinical symptoms and improve behavior. With the aid of semantically enabled informatics portals, we are able to understand different types of recovery patterns, the underlying mechanisms of treatments, and how the symptoms interact with each other.
Neural network modeling has shown great potential in the field of unipolar depression study. In order to get a good sense of the parameters, most neural network models need a large quantity of data. One of the big issues in neural network modeling is how to search and pre-process raw data, transform them into appropriate formats, implement data integration, acquire knowledge from different computational models, and ensure the high quality of data. Semantic web technology helps to address these difficulties.
In this study, we mainly present a semantic technology-based approach for building the information portal for major depressive disorders. The exemplar portal captures the semantics of domain knowledge using a family of strongly related medical ontologies we designed. Such as the Hamilton scales, depressive patients and psychiatrists. Next, through the help of Google Refine, the informatics portal integrates patient's clinical data from several different sources with various formats following the linked data principles. We transform the medical data into a RDF and store them into sematic triple store: AllegroGraph. In this way, data can be easily retrieved via the JavaScript package JQuery and query language SPARQL.
August 2012
School of Engineering
Dept. of Industrial and Systems Engineering
Rensselaer Polytechnic Institute, Troy, NY
Rensselaer Theses and Dissertations Online Collection
CC BY-NC-ND. Users may download and share copies with attribution in accordance with a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License. No commercial use or derivatives are permitted without the explicit approval of the author.