Search

At public sex environments in four U.S. cities, 1,369 men who have sex with men (MSM) were asked about sexual self-identification, recent HIV risk behaviors, and exposures to HIV information. Half of respondents (n = 687) self-identified as gay, 40% (n = 546) as bisexual, and 10% (n = 136) as straight. Ninety-nine percent of both gay and bisexual MSM and 96% of straight MSM reported oral sex with men; 94%, 68%, and 46%, respectively, reported anal sex with men, while 62%, 98%, and 97%, respectively, reported vaginal sex with women. Recent exposure to any HIV information was reported by 96%, 91%, and 89% respectively of gay, bisexual, and straight MSM; gay MSM were most likely to get HIV information from talking with someone. However, television was the only medium to reach more than half of gay, bisexual, and straight MSM. Non-gay-identified MSM and their partners are at high risk for HIV transmission, but more study is needed to identify the most effective channels for conveying risk reduction messages to this population.

Active learning differs from “learning from examples” in that the learning algorithm assumes at least some control over what part of the input domain it receives information about. In some situations, active learning is provably more powerful than learning from examples alone, giving better generalization for a fixed number of training examples. In this article, we consider the problem of learning a binary concept in the absence of noise. We describe a formalism for active concept learning called selective sampling and show how it may be approximately implemented by a neural network. In selective sampling, a learner receives distribution information from the environment and queries an oracle on parts of the domain it considers “useful.” We test our implementation, called an SG-network, on three domains and observe significant improvement in generalization.