Print Email Facebook Twitter Stability-Based Model Selection in Non-Stationary Environment Title Stability-Based Model Selection in Non-Stationary Environment Author Ivanov, Viktor (TU Delft Electrical Engineering, Mathematics and Computer Science) Contributor Tax, David (mentor) Degree granting institution Delft University of Technology Date 2017-08-31 Abstract Model selection is associated to model assessment, which is the problem of comparing different models, or model hyperparameters, for a particular learning task. It constitutes a fundamental step in building machine learning models. The central question is: How a model will work in the future? In this thesis, a new model selection scheme for learning algorithms operating in non-stationary environment is introduced which is based on a notion of stability. In the evaluation part, the performance of the stability measure is studied for loss given default (LGD) model selection. The evaluation is based on a real-world LGD dataset provided by an anonymous financial institution. Subject machine learningconcept driftmodel validationcredit riskloss given default To reference this document use: http://resolver.tudelft.nl/uuid:31bbf0f5-c43a-4dc0-b2d5-59880368f2a3 Part of collection Student theses Document type master thesis Rights © 2017 Viktor Ivanov Files PDF viktor_ivanov_msc_thesis.pdf 998.82 KB Close viewer /islandora/object/uuid:31bbf0f5-c43a-4dc0-b2d5-59880368f2a3/datastream/OBJ/view