Learning to Score
This addresses a problem in machine learning for scenarios like healthcare where disease severity scoring is needed without well-defined criteria, but it appears incremental as it builds on existing concepts like side information and metric learning.
The paper tackles the problem of learning from data where target labels are unavailable but side information is available, by formulating an ensemble of representation learning, side information, and metric learning to create a scoring model. It demonstrates utility on benchmark datasets and biomedical patient records, though no concrete numbers are provided.
Common machine learning settings range from supervised tasks, where accurately labeled data is accessible, through semi-supervised and weakly-supervised tasks, where target labels are scant or noisy, to unsupervised tasks where labels are unobtainable. In this paper we study a scenario where the target labels are not available but additional related information is at hand. This information, referred to as Side Information, is either correlated with the unknown labels or imposes constraints on the feature space. We formulate the problem as an ensemble of three semantic components: representation learning, side information and metric learning. The proposed scoring model is advantageous for multiple use-cases. For example, in the healthcare domain it can be used to create a severity score for diseases where the symptoms are known but the criteria for the disease progression are not well defined. We demonstrate the utility of the suggested scoring system on well-known benchmark data-sets and bio-medical patient records.