Predicting Contextual Sequences via Submodular Function Maximization

TL;DR

This paper introduces a context-aware sequence optimization method using submodular function maximization, improving task-specific ordering in applications like robotics by learning classifiers for sequence slots.

Contribution

It presents a novel reduction-based approach that incorporates context into sequence optimization, leveraging submodular maximization for improved decision-making.

Findings

Effective in robotics control tasks

Outperforms static ordering methods

Demonstrates improved trajectory and path planning

Abstract

Sequence optimization, where the items in a list are ordered to maximize some reward has many applications such as web advertisement placement, search, and control libraries in robotics. Previous work in sequence optimization produces a static ordering that does not take any features of the item or context of the problem into account. In this work, we propose a general approach to order the items within the sequence based on the context (e.g., perceptual information, environment description, and goals). We take a simple, efficient, reduction-based approach where the choice and order of the items is established by repeatedly learning simple classifiers or regressors for each "slot" in the sequence. Our approach leverages recent work on submodular function maximization to provide a formal regret reduction from submodular sequence optimization to simple cost-sensitive prediction. We apply our contextual sequence prediction algorithm to optimize control libraries and demonstrate results on two robotics problems: manipulator trajectory prediction and mobile robot path planning.