Assessing AI systems is difficult. Humans rely on AI systems in increasing ways, both visible and invisible, meaning a variety of stakeholders need a variety of assessment tools (e.g., a professional auditor, a developer, and an end user all have different needs). We posit that it is possible to provide...
Ecological domains seeking to understand the environment and the behavior of species have received little attention in machine learning (ML), despite the fact that environmental changes have a significant impact on humans as well as ecosystems. Some ecological problems can be formulated similarly to other common ML applications, but there...
Collective robotic systems are biologically-inspired and exhibit behaviors found in spatial swarms (e.g., fish), colonies (e.g., ants), or a combination of both (e.g., bees). Collective robotic system popularity continues to increase due to their apparent global intelligence and emergent behaviors. Many applications can benefit from the incorporation of collectives, including...
Maintaining the sustainability of the earth’s ecosystems has attracted much attention as these ecosystems are facing more and more pressure from human activities. Machine learning can play an important role in promoting sustainability as a large amount of data is being collected from ecosystems. There are at least three important...
How can end users efficiently influence the predictions that machine learning systems make on their behalf? Traditional systems rely on users to provide examples of how they want the learning system to behave, but this is not always practical for the user, nor efficient for the learning system. This dissertation...
This dissertation explores algorithms for learning ranking functions to efficiently solve search problems, with application to automated planning. Specifically, we consider the frameworks of beam search, greedy search, and randomized search, which all aim to maintain tractability at the cost of not guaranteeing completeness nor optimality. Our learning objective for...
In text classification, labeling features is often less time consuming than labeling entire documents. In situations where very little labeled training data is available, feature relevance feedback has the potential to dramatically increase classification performance. We review previous work on incorporating feature relevance feedback in the form of labeled features...
End users' programs are fraught with errors, costing companies millions of dollars. One reason may be that researchers and tool designers have not yet focused on end-user debugging strategies. To investigate this possibility, this dissertation presents eight empirical studies and a new strategy-based end-user debugging tool for Excel, called StratCel....