Aggregated f-average Neural Network applied to Few-Shot Class Incremental Learning
This work addresses incremental learning with limited data, but it appears incremental as it fuses existing ensembling frameworks without claiming broad SOTA or foundational impact.
The paper tackles the problem of few-shot class incremental learning by introducing an aggregated f-average shallow neural network that models and combines different types of averages to optimally aggregate weak learners' predictions, achieving good performance as illustrated in the results.
Ensemble learning leverages multiple models (i.e., weak learners) on a common machine learning task to enhance prediction performance. Basic ensembling approaches average the weak learners outputs, while more sophisticated ones stack a machine learning model in between the weak learners outputs and the final prediction. This work fuses both aforementioned frameworks. We introduce an aggregated f-average (AFA) shallow neural network which models and combines different types of averages to perform an optimal aggregation of the weak learners predictions. We emphasise its interpretable architecture and simple training strategy, and illustrate its good performance on the problem of few-shot class incremental learning.