| Topic: |
Principal component analysis based unsupervised feature extraction applied to bioinformatics analysis |
| Speaker: |
Professor Y-h. Taguchi ( Department of Physics, Chuo University, Japan) |
| Time: |
Fri. Oct. 28, 2016, 10:40 - 12:00 |
| Place: |
GenIII R837 |
| Abstract: |
In bioinformatics analysis, it is very usual that features (genes) are more than samples. In this case, reduction of features is critical to analyze data. Although there were many complicated procedures proposed, recently we found that principal component analysis (PCA) can be successfully used for this purpose. Featured placed as outliers in the embedding turn out to be critical for bioinfotmatics analysis. We will show some of recent results using this methodology including miRNA-mRNA interaction from expression analysis. |