Balancing Exploration and Exploitation: A New Algorithm for Active Machine Learning

Active machine learning algorithms are used when large numbers of unlabeled examples are available and getting labels for them is costly (e.g. requiring consulting a human expert). Many conventional active learning algorithms focus on refining the decision boundary, at the expense of exploring new regions that the current hypothesis misclassifies. We propose a new active learning algorithm that balances such exploration with refining of the decision boundary by dynamically adjusting the probability to explore at each step. Our experimental results demonstrate improved performance on data sets that require extensive exploration while remaining competitive on data sets that do not. Our algorithm also shows significant tolerance of noise.

Deer Guards and Bump Gates for Excluding White-Tailed Deer From Fenced Resources

White-tailed deer (Odocoileus virginianus) causing damage is a reoccurring theme in the realm of wildlife damage management, especially regarding human safety, disease transmission, and agricultural losses. Fences often are the only reliable long-term nonlethal means of controlling deer damage. The efficacy of fences, however, relies on their weakest link: human-operated gates. Although not overly time-consuming, the act of closing a gate appears to be a burden to individuals, resulting in open-access to an otherwise protected resource. We examined the efficacy of 2 alternatives to traditional gates to evaluate their potential to be used for excluding or containing deer. We evaluated a commercially available kit for mechanically opening and closing gates and a modified deer guard that resembles a common cattle guard but incorporates bearing-mounted rollers as cross members. The gate kit proved effective in restricting deer access to bait throughout the study, but, in supplemental evaluations, we observed excessive rates of functional failure. Deer guards reduced deer entry into exclosures, but efficacy declined with time as deer walked and jumped across guards. With some refining, both guards and gates have potential to be useful components of an integrated biosecurity strategy.