Loss Functions for Multiset Prediction
This work addresses a challenging problem in supervised learning for applications requiring unordered, repetitive predictions, but it is incremental as it builds on existing loss function approaches.
The paper tackles the problem of multiset prediction, where the goal is to predict a multiset of items without a known order and with possible repetitions, by proposing a novel loss function based on sequential decision making. The experiments on synthetic and real datasets show that the proposed loss function is more effective than baseline methods like reinforcement learning and sequence loss functions.
We study the problem of multiset prediction. The goal of multiset prediction is to train a predictor that maps an input to a multiset consisting of multiple items. Unlike existing problems in supervised learning, such as classification, ranking and sequence generation, there is no known order among items in a target multiset, and each item in the multiset may appear more than once, making this problem extremely challenging. In this paper, we propose a novel multiset loss function by viewing this problem from the perspective of sequential decision making. The proposed multiset loss function is empirically evaluated on two families of datasets, one synthetic and the other real, with varying levels of difficulty, against various baseline loss functions including reinforcement learning, sequence, and aggregated distribution matching loss functions. The experiments reveal the effectiveness of the proposed loss function over the others.