POPhillip Odom

Phillip Odom

*This is an algorithmically generated author record based on publicly available information.

Phillip Odom has a total of 10 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Human-Allied Learning of Probabilistic Models from Relational Data

    by Harsha Kokel, Nandini Ramanan, Phillip Odom, Erik P. Blasch, Sriraam Natarajan

  • Open Access

    Barrier Function Overrides For Non-Convex Fixed Wing Flight Control and Self-Driving Cars

    by Eric Squires, Phillip Odom, Zsolt Kira

    Reinforcement Learning (RL) has enabled vast performance improvements for robotics systems. To achieve these results though, the agent often must randomly explore the environment, which for safety critical systems presents a significant challenge. Barrier functions can solve this challenge by enabling an override that approximates the RL control input as closely as possible without violating a safety constraint. Unfortunately, this override can be computationally intractable in cases where the dynamics are not convex in the control input or when time is discrete, as is often the case when training RL systems. We therefore consider these cases, developing novel barrier functions for two non-convex systems (fixed wing aircraft and self-driving cars performing lane merging with adaptive cruise control) in discrete time. Although solving for an online and optimal override is in general intractable when the dynamics are nonconvex in the control input, we investigate approximate solutions, finding that these approximations enable performance commensurate with baseline RL methods with zero safety violations. In particular, even without attempting to solve for the optimal override at all, performance is still competitive with baseline RL performance. We discuss the tradeoffs of the approximate override solutions including performance and computational tractability.

  • Open Access

    Knowledge-based Refinement of Scientific Publication Knowledge Graphs

    by Siwen Yan, Phillip Odom, Sriraam Natarajan

    We consider the problem of identifying authorship by posing it as a knowledge graph construction and refinement. To this effect, we model this problem as learning a probabilistic logic model in the presence of human guidance (knowledge-based learning). Specifically, we learn relational regression trees using functional gradient boosting that outputs explainable rules. To incorporate human knowledge, advice in the form of first-order clauses is injected to refine the trees. We demonstrate the usefulness of human knowledge both quantitatively and qualitatively in seven authorship domains.

  • Open Access

    Multi-class Classification without Multi-class Labels

    by Yen-Chang Hsu, Zhaoyang Lv, Joel Schlosser, Phillip Odom, Zsolt Kira

    This work presents a new strategy for multi-class classification that requires no class-specific labels, but instead leverages pairwise similarity between examples, which is a weaker form of annotation. The proposed method, meta classification learning, optimizes a binary classifier for pairwise similarity prediction and through this process learns a multi-class classifier as a submodule. We formulate this approach, present a probabilistic graphical model for it, and derive a surprisingly simple loss function that can be used to learn neural network-based models. We then demonstrate that this same framework generalizes to the supervised, unsupervised cross-task, and semi-supervised settings. Our method is evaluated against state of the art in all three learning paradigms and shows a superior or comparable accuracy, providing evidence that learning multi-class classification without multi-class labels is a viable learning option.