Search for: All records

Abstract We present a composite machine learning framework to estimate posterior probability distributions of bulge-to-total light ratio, half-light radius, and flux for active galactic nucleus (AGN) host galaxies withinz < 1.4 andm < 23 in the Hyper Supreme-Cam (HSC) Wide survey. We divide the data into five redshift bins:low(0 < z < 0.25),mid(0.25 < z < 0.5),high(0.5 < z < 0.9),extra(0.9 < z < 1.1), andextreme(1.1 < z < 1.4), and train our models independently in each bin. We use PSFGAN to decompose the AGN point-source light from its host galaxy, and invoke the Galaxy Morphology Posterior Estimation Network (GaMPEN) to estimate morphological parameters of the recovered host galaxy. We first trained our models on simulated data, and then fine-tuned our algorithm via transfer learning using labeled real data. To create training labels for transfer learning, we used GALFIT to fit  ∼20,000 real HSC galaxies in each redshift bin. We comprehensively examined that the predicted values from our final models agree well with the GALFIT values for the vast majority of cases. Our PSFGAN + GaMPEN framework runs at least three orders of magnitude faster than traditional light-profile fitting methods, and can be easily retrained for other morphological parameters or on other data sets with diverse ranges of resolutions, seeing conditions, and signal-to-noise ratios, making it an ideal tool for analyzing AGN host galaxies from large surveys coming soon from the Rubin-LSST, Euclid, and Roman telescopes. 

Abstract Background Humans emit approximately 30 million microbial cells per hour into their immediate vicinity. However, sampling of aerosolized microbial taxa (aerobiome) remains largely uncharacterized due to the complexity and limitations of sampling techniques, which are highly susceptible to low biomass and rapid sample degradation. Recently, there has been an interest in developing technology that collects naturally occurring water from the atmosphere, even within the built environment. Here, we analyze the feasibility of indoor aerosol condensation collection as a method to capture and analyze the aerobiome. Methods Aerosols were collected via condensation or active impingement in a laboratory setting over the course of 8 h. Microbial DNA was extracted from collected samples and sequenced (16S rRNA) to analyze microbial diversity and community composition. Dimensional reduction and multivariate statistics were employed to identify significant ( p  < 0.05) differences in relative abundances of specific microbial taxa observed between the two sampling platforms. Results Aerosol condensation capture is highly efficient with a yield greater than 95% when compared to expected values. Compared to air impingement, aerosol condensation showed no significant difference (ANOVA, p  > 0.05) in microbial diversity. Among identified taxa, Streptophyta and Pseudomonadales comprised approximately 70% of the microbial community composition. Conclusion The results suggest that condensation of atmospheric humidity is a suitable method for the capture of airborne microbial taxa reflected by microbial community similarity between devices. Future investigation of aerosol condensation may provide insight into the efficacy and viability of this new tool to investigate airborne microorganisms. Importance On average, humans shed approximately 30 million microbial cells each hour into their immediate environment making humans the primary contributor to shaping the microbiome found within the built environment. In addition, recent events have highlighted the importance of understanding how microorganisms within the built environment are aerosolized and dispersed, but more importantly, the lack in development of technology that is capable of actively sampling the ever-changing aerosolized microbiome, i.e., aerobiome. This research highlights the capability of sampling the aerobiome by taking advantage of naturally occurring atmospheric humidity. Our novel approach reproduces the biological content in the atmosphere and can provide insight into the environmental microbiology of indoor spaces. 

Feature acquisition in predictive modeling is an important task in many practical applications. For example, in patient health prediction, we do not fully observe their personal features and need to dynamically select features to acquire. Our goal is to acquire a small subset of features that maximize prediction performance. Recently, some works reformulated feature acquisition as a Markov decision process and applied reinforcement learning (RL) algorithms, where the reward reflects both prediction performance and feature acquisition cost. However, RL algorithms only use zeroth-order information on the reward, which leads to slow empirical convergence, especially when there are many actions (number of features) to consider. For predictive modeling, it is possible to use first-order information on the reward, i.e., gradients, since we are often given an already collected dataset. Therefore, we propose differentiable feature acquisition (DiFA), which uses a differentiable representation of the feature selection policy to enable gradients to flow from the prediction loss to the policy parameters. We conduct extensive experiments on various real-world datasets and show that DiFA significantly outperforms existing feature acquisition methods when the number of features is large. 

In sequential recommender system applications, it is important to develop models that can capture users' evolving interest over time to successfully recommend future items that they are likely to interact with. For users with long histories, typical models based on recurrent neural networks tend to forget important items in the distant past. Recent works have shown that storing a small sketch of past items can improve sequential recommendation tasks. However, these works all rely on static sketching policies, i.e., heuristics to select items to keep in the sketch, which are not necessarily optimal and cannot improve over time with more training data. In this paper, we propose a differentiable policy for sketching (DiPS), a framework that learns a data-driven sketching policy in an end-to-end manner together with the recommender system model to explicitly maximize recommendation quality in the future. We also propose an approximate estimator of the gradient for optimizing the sketching algorithm parameters that is computationally efficient. We verify the effectiveness of DiPS on real-world datasets under various practical settings and show that it requires up to 50% fewer sketch items to reach the same predictive quality than existing sketching policies.