Weakly Supervised Label Learning Flows
This addresses the need for cheaper labeling in machine learning tasks, but it is incremental as it builds on existing weakly supervised methods with a novel generative approach.
The paper tackles the problem of costly ground-truth labeling in supervised learning by proposing label learning flows (LLF), a generative framework using normalizing flows to optimize conditional likelihoods of labelings constrained by weak signals, and it shows that LLF outperforms baselines in experiments on three weakly supervised learning problems.
Supervised learning usually requires a large amount of labelled data. However, attaining ground-truth labels is costly for many tasks. Alternatively, weakly supervised methods learn with cheap weak signals that only approximately label some data. Many existing weakly supervised learning methods learn a deterministic function that estimates labels given the input data and weak signals. In this paper, we develop label learning flows (LLF), a general framework for weakly supervised learning problems. Our method is a generative model based on normalizing flows. The main idea of LLF is to optimize the conditional likelihoods of all possible labelings of the data within a constrained space defined by weak signals. We develop a training method for LLF that trains the conditional flow inversely and avoids estimating the labels. Once a model is trained, we can make predictions with a sampling algorithm. We apply LLF to three weakly supervised learning problems. Experiment results show that our method outperforms many baselines we compare against.