Automatic gender recognition (AGR) is a subfield of facial recognition that has recently been scrutinized for bias in the form of misgendering and erasure against various identity groups in our society. Recent studies have found that several commercial AGR classifiers (from Microsoft, IMB, Face++) are biased against women and darker-skinned people as well as gender non-binary people. In this work, we investigate and quantify AGR classifier bias against transgender people by developing and evaluating three different convolutional neural networks (CNN): using images of cisgender individuals, using images of transgender individuals, and using images of both cisgender and transgender individuals. We find that the cisgender trained classifier is 91.7% accurate when evaluated on cisgender people, but only 68.9% accurate when evaluated on transgender people, with the worst performance on trans men with 38.6% precision. We also find that the classifier trained on the combined dataset performs nearly as well as and occasionally outperforms both other classifiers when evaluated on their own datasets, highlighting potential methods for avoiding overfitting. Additionally, we visualize how the classifiers differ by obscuring different parts of the face. Overall, the disparities of accuracy between each classifier demonstrate the degree to which they are impacted by the composition on their dataset and highlight the possibility for commercial AGR classifiers potential to misgender trans people, in particular, transgender men, at a high rate.

Debris Disk Stars are stars that are surrounded by dust and diffuse gas. Certain elements in the gas and dust absorb the light from the star and show up as an absorption-line spectrum. Our goal was to find trends in the elemental abundance patterns of similar debris disk stars. We took 30 debris disk stars that are similar in mass and surface temperature and measured their elemental abundances. Using Spectroscopy through IRAF’s s-plot tool we found accurate abundances of certain elements. We found abundances of metals including Iron, Nickel, Sodium, Carbon, and Oxygen. Comparing their Metallicities ([x/Fe]) we can find certain trends between the abundance of the elements and their condensation temperatures. From the 30 stars in the data-set, we chose 3 of the most interesting objects. The first star we chose, HD162826, is the closest in terms of motion and chemical abundances to the Sun. The second star, HD187691, has a small debris disk. The third star, HD122652, has a large debris disk. For HD162826, we found the metallicity of the refractory elements (high condensation temperature, >1200 K) to be spread around the 0 metallicity marker ([x/Fe]=0), for HD187691 we found most of the refractory elements above [x/Fe]=0 with the rest being very close to zero, and for HD122652 we found that the metallicities of most of the refractory elements were negative. The trends we found are small and could be explained by observational uncertainty, therefore further analysis of our data set would be required to make stronger conclusions. By accurately measuring more absorption lines from our dataset, possibly more connections can be made about the properties of Debris Disk Stars.

In 2007, David Vokoun et al. derived a formula for the force of interaction between magnets. The formula is called the Gilbert's Model. According to the Gilbert’s Model, the force between two ferromagnets is given by a constant factor proportional to the saturation magnetization of each magnet multiplied by a function of the separation distance and geometry of the magnets. We show that the assumed constant is better described as a function of hyperbolic tangent of the separation distance due to the effects of magnetic field interactions on the magnetizations of each magnet, and we demonstrate that the inclusion of a simple toy 1D Ising model acting as a perturbation on the background magnetizations better predicts magnetic coupling of cylindrical magnets over small distances.