Investigation of image processing algorithms for medical application
Gene Regulatory Network is the network that constitute the interaction between genes. There is a need to find an effective way for the discovery of gene regulatory network which will provide better information for further advancement in biotechnology and bioinformatics. One of the prominent approach...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/63880 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Gene Regulatory Network is the network that constitute the interaction between genes. There is a need to find an effective way for the discovery of gene regulatory network which will provide better information for further advancement in biotechnology and bioinformatics. One of the prominent approach to analyse GRN is Granger Causality Analysis. This project tries to perform comparative study between different implementation of granger causality. Specifically Multivariate Granger Causality (MVGC), Lasso Granger Causality, and Copula Granger Causality. These three methods are previously researched by other researcher, but never been compared side to side under the same condition. The project was implemented with experimental framework that utilizes 3 control variables and 7 metrics. These 7 metrics are the basis for the comparative study of the project itself. Based on the findings in the experiment, overall score favours MVGC methods as the best algorithm. However in some condition, Lasso and Copula score exceeds MVGC which indicates that the results are conditional, not absolute. It could be said that each method has the condition that maximizes their performance. This project provides analysis of these situations based on the experiment result. |
|---|