Semi-supervised learning of sparse linear models in mass spectral imaging

摘要: We present an approach to learn predictive models and perform variable selection by incorporating structural information from Mass Spectral Imaging (MSI) data. explore the use of a smooth quadratic penalty model natural ordering physical variables, that is mass-to-charge (m/z) ratios. Thereby, estimated parameters for nearby variables are enforced smoothly vary. Similarly, overcome lack labeled data we spatial proximity among spectra means connectivity graph over set predicted labels. usefulness this in mouse brain MSI set.