Estimating Missing Values in Gene Expression Data Using Support Vector Regression
Author Information
Author(s): Wang Xian, Li Ao, Jiang Zhaohui, Feng Huanqing
Primary Institution: University of Science and Technology of China
Hypothesis
Can Support Vector Regression (SVR) effectively estimate missing values in gene expression data?
Conclusion
The SVR impute method shows better performance than previously developed methods for estimating missing values in gene expression data.
Supporting Evidence
- The SVR method outperformed KNN, BPCA, and LLS impute methods in estimating missing values.
- The orthogonal input coding scheme provided the best estimation results with the minimum Normalized Root Mean Squared Error (NRMSE).
- The SVR method demonstrated robust performance across various datasets with different missing value percentages.
Takeaway
This study found a way to fill in missing information in gene data using a smart math method called Support Vector Regression, which helps scientists understand genes better.
Methodology
The study used Support Vector Regression with an orthogonal coding scheme to estimate missing values in six gene expression datasets.
Limitations
The performance of the SVR method may decrease when the percentage of missing values is too high.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website