Abstract-This paper considers fitting a mixture of Gaussians model to high-dimensional data in scenarios where there are fewer data samples than feature dimensions. Issues that arise when using principal component analysis (PCA) to represent Gaussian distributions inside Expectation-Maximization (EM) are addressed, and a practical algorithm results. Unlike other algorithms that have been proposed, this algorithm does not try to compress the data to fit low-dimensional models. Instead, it models Gaussian distributions in the ( 1)-dimensional space spanned by the data samples. We are able to show that this algorithm converges on data sets where low-dimensional techniques do not.
This study evaluated cataracts in wild boar exposed to chronic low-dose radiation. We examined wild boar from within and outside the Fukushima Exclusion Zone for nuclear, cortical, and posterior subcapsular (PSC) cataracts in vivo and photographically. Plausible upper-bound, lifetime radiation dose for each boar was estimated from radioactivity levels in each animal's home range combined with tissue concentrations of 134+137 Cesium. Fifteen exposed and twenty control boar were evaluated. There were no significant differences in overall prevalence or score for cortical or PSC cataracts between exposed and control animals. Nuclear (centrally located) cataracts were significantly more prevalent in exposed boar (p < 0.05) and had statistically higher median scores. Plausible upper-bound, lifetime radiation dose ranged from 1 to 1,600 mGy in exposed animals, with no correlation between dose and cortical or PSC score. While radiation dose and nuclear score were positively associated, the impact of age could not be completely separated from the relationship. Additionally, the clinical significance of even the highest scoring nuclear cataract was negligible. Based on the population sampled, wild boar in the Fukushima Exclusion Zone do not have a significantly higher prevalence or risk of cortical or PSC cataracts compared to control animals.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.