One-Shot Induction of Generalized Logical Concepts via Human Guidance
This addresses the challenge of sample-efficient concept induction in AI, with incremental improvements for logic-based learning systems.
The paper tackles the problem of learning generalized first-order concept representations from a single example by augmenting an inductive logic programming learner with a novel distance measure and human advice, resulting in improved effectiveness and efficiency over existing methods in diverse tasks.
We consider the problem of learning generalized first-order representations of concepts from a single example. To address this challenging problem, we augment an inductive logic programming learner with two novel algorithmic contributions. First, we define a distance measure between candidate concept representations that improves the efficiency of search for target concept and generalization. Second, we leverage richer human inputs in the form of advice to improve the sample-efficiency of learning. We prove that the proposed distance measure is semantically valid and use that to derive a PAC bound. Our experimental analysis on diverse concept learning tasks demonstrates both the effectiveness and efficiency of the proposed approach over a first-order concept learner using only examples.