Showing 1–1 of 1 results for author: Sivrikaya, O E

Learning Maximally Predictive Prototypes in Multiple Instance Learning

Authors: Mert Yuksekgonul, Ozgur Emre Sivrikaya, Mustafa Gokce Baydogan

Abstract: In this work, we propose a simple model that provides permutation invariant maximally predictive prototype generator from a given dataset, which leads to interpretability of the solution and concrete insights to the nature and the solution of a problem. Our aim is to find out prototypes in the feature space to map the collection of instances (i.e. bags) to a distance feature space and simultaneous… ▽ More In this work, we propose a simple model that provides permutation invariant maximally predictive prototype generator from a given dataset, which leads to interpretability of the solution and concrete insights to the nature and the solution of a problem. Our aim is to find out prototypes in the feature space to map the collection of instances (i.e. bags) to a distance feature space and simultaneously learn a linear classifier for multiple instance learning (MIL). Our experiments on classical MIL benchmark datasets demonstrate that proposed framework is an accurate and efficient classifier compared to the existing approaches. △ Less

Submitted 22 January, 2021; v1 submitted 2 October, 2019; originally announced October 2019.

Comments: Sets & Partitions Workshop at NeurIPS 2019