Developing Model Strings for Model Managers

Abstract

:Models are a key component in the decision-making process. Model managers have been proposed to select and direct the use of models in a decision setting. A limitation of model managers is their inability to optimally integrate several models to arrive at a decision. A scheme to overcome this limitation is presented. By specifying input requirements to each model and outputs available from each model, a string of models may be developed. An implicit enumeration algorithm ensures a feasible string of models for a required output and composes the string in optimal form according to a user-defined objective. The algorithm may be used in conjunction with existing model management techniques.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 21 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.