Human-Guided Learning of Column Networks: Augmenting Deep Learning with Advice
This addresses a domain-specific challenge in deep learning for structured domains with limited data, representing an incremental improvement.
The paper tackles the problem of sub-optimal learning in Column Networks due to sparse and noisy samples by augmenting them with human advice, resulting in either superior overall performance or faster convergence.
Recently, deep models have been successfully applied in several applications, especially with low-level representations. However, sparse, noisy samples and structured domains (with multiple objects and interactions) are some of the open challenges in most deep models. Column Networks, a deep architecture, can succinctly capture such domain structure and interactions, but may still be prone to sub-optimal learning from sparse and noisy samples. Inspired by the success of human-advice guided learning in AI, especially in data-scarce domains, we propose Knowledge-augmented Column Networks that leverage human advice/knowledge for better learning with noisy/sparse samples. Our experiments demonstrate that our approach leads to either superior overall performance or faster convergence (i.e., both effective and efficient).