Certainty Pooling for Multiple Instance Learning
This work addresses the challenge of improving prediction accuracy and interpretability in weakly supervised learning for domains like histopathology, though it is incremental as it builds on existing pooling methods.
The paper tackles the problem of robust and explainable bag-level predictions in Multiple Instance Learning by introducing Certainty Pooling, which incorporates model certainty into pooling operators, resulting in outperformance in bag-level and instance-level predictions on MNIST and Camelyon16 datasets, especially with small training sets.
Multiple Instance Learning is a form of weakly supervised learning in which the data is arranged in sets of instances called bags with one label assigned per bag. The bag level class prediction is derived from the multiple instances through application of a permutation invariant pooling operator on instance predictions or embeddings. We present a novel pooling operator called \textbf{Certainty Pooling} which incorporates the model certainty into bag predictions resulting in a more robust and explainable model. We compare our proposed method with other pooling operators in controlled experiments with low evidence ratio bags based on MNIST, as well as on a real life histopathology dataset - Camelyon16. Our method outperforms other methods in both bag level and instance level prediction, especially when only small training sets are available. We discuss the rationale behind our approach and the reasons for its superiority for these types of datasets.