부산대학교 도서관

Transfer learning techniques use information from pre-trained machine learning models (e.g., information on the feature relevance or feature effects with respect to a clinical outcome of interest) to apply it to a new but similar data analysis task, in order to exploit the prior information to build more robust and accurate models (see [[150]] for a review of methodologies). Tailored software solutions have been made available to preprocess clinical data [[21]], NGS data [[49]], microarray data [[50]], different types of metabolomics and proteomics data [[18]], and cellular and brain imaging data [[51]-[54]]. Relevant quality controls typically include statistical outlier checks and computing data type-specific quality metrics, as implemented in established software packages, e.g., the I fastQC/FQC i package for next-generation sequencing (NGS) data [[14]], I arrayQualityMetrics i for microarray data [[15]], I pseudoQC i , I MeTaQuaC i , and I Normalyzer i for proteomics and metabolomics data [[16]-[18]]. [Extracted from the article]