Temporally adaptive estimation of logistic classifiers on data streams

Modern technology has allowed real-time data collection in a variety of domains, ranging from environmental monitoring to healthcare. Consequently, there is a growing need for algorithms capable of performing inferential tasks in an online manner, continuously revising their estimates to reflect the current status of the underlying process. In particular, we are interested in constructing online and temporally adaptive classifiers capable of handling the possibly drifting decision boundaries arising in streaming environments. We first make a quadratic approximation to the log-likelihood that yields a recursive algorithm for fitting logistic regression online. We then suggest a novel way of equipping this framework with self-tuning forgetting factors. The resulting scheme is capable of tracking changes in the underlying probability distribution, adapting the decision boundary appropriately and hence maintaining high classification accuracy in dynamic or unstable environments. We demonstrate the scheme's effectiveness in both real and simulated streaming environments. © Springer-Verlag 2009.

Do we know the strength of the chorioamnion? A critical review and analysis

The chorioamnion is the membrane that surrounds the fetus during gestation. Normally, it must remain intact for the duration of pregnancy, 37-42 weeks, and only rupture during or just before labour and delivery of the fetus. In a significant number (3%) of all births, this does not happen, and membranes rupture before term, resulting in preterm birth and significant perinatal morbidity. It is known that the material properties of chorioamnion may play a major role in mechanical rupture; a number of studies have been undertaken to characterise the physical nature of chorioamnion and examine factors that may predispose to rupture. However, the existing literature is inconsistent in its choice of both physical testing methods and data analysis techniques, motivating the current review. Experimental data from a large number of chorioamnion mechanical studies were collated, and data were converted to standard engineering quantities. The failure strength of the chorioamnion membrane was found consistently to value approximately 0.9 MPa. It is hoped that past and future studies of membrane mechanics can provide insight into the role of chorioamnion in labour and delivery. In addition, biomechanical approaches can help elucidate the potential causes of early rupture, and suggest future protocols or treatments that could both diagnose and prevent its occurrence. © 2009 Elsevier Ireland Ltd.