Sequential feature selection for classification

Abstract

In most real-world information processing problems, data is not a free resource; its acquisition is rather time-consuming and/or expensive. We investigate how these two factors can be included in supervised classication tasks by deriving classication as a sequential decision process and making it accessible to Reinforcement Learning. Our method performs a sequential feature selection that learns which features are most informative at each timestep, choosing the next feature depending on the already selected features and the internal belief of the classier. Experiments on a handwritten digits classication task show signicant reduction in required data for correct classication, while a medical diabetes prediction task illustrates variable feature cost minimization as a further property of our algorithm.