Source author record

Edward J. Kim

Edward J. Kim appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

astro-ph.CO astro-ph.IM gr-qc hep-th astro-ph.GA Computer Vision cs.CY physics.ed-ph stat.OT

Catalog footprint

What is connected

5works

9topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2016arXiv

Star-galaxy Classification Using Deep Convolutional Neural Networks

Most existing star-galaxy classifiers use the reduced summary information from catalogs, requiring careful feature extraction and selection. The latest advances in machine learning that use deep convolutional neural networks allow a machine to automatically learn the features directly from data, minimizing the need for input from human experts. We present a star-galaxy classification framework that uses deep convolutional neural networks (ConvNets) directly on the reduced, calibrated pixel values. Using data from the Sloan Digital Sky Survey (SDSS) and the Canada-France-Hawaii Telescope Lensing Survey (CFHTLenS), we demonstrate that ConvNets are able to produce accurate and well-calibrated probabilistic classifications that are competitive with conventional machine learning techniques. Future advances in deep learning may bring more success with current and forthcoming photometric surveys, such as the Dark Energy Survey (DES) and the Large Synoptic Survey Telescope (LSST), because deep neural networks require very little, manual feature engineering.

preprint2016arXiv

Teaching Data Science

We describe an introductory data science course, entitled Introduction to Data Science, offered at the University of Illinois at Urbana-Champaign. The course introduced general programming concepts by using the Python programming language with an emphasis on data preparation, processing, and presentation. The course had no prerequisites, and students were not expected to have any programming experience. This introductory course was designed to cover a wide range of topics, from the nature of data, to storage, to visualization, to probability and statistical analysis, to cloud and high performance computing, without becoming overly focused on any one subject. We conclude this article with a discussion of lessons learned and our plans to develop new data science courses.

preprint2015arXiv

A Hybrid Ensemble Learning Approach to Star-Galaxy Classification

There exist a variety of star-galaxy classification techniques, each with their own strengths and weaknesses. In this paper, we present a novel meta-classification framework that combines and fully exploits different techniques to produce a more robust star-galaxy classification. To demonstrate this hybrid, ensemble approach, we combine a purely morphological classifier, a supervised machine learning method based on random forest, an unsupervised machine learning method based on self-organizing maps, and a hierarchical Bayesian template fitting method. Using data from the CFHTLenS survey, we consider different scenarios: when a high-quality training set is available with spectroscopic labels from DEEP2, SDSS, VIPERS, and VVDS, and when the demographics of sources in a low-quality training set do not match the demographics of objects in the test data set. We demonstrate that our Bayesian combination technique improves the overall performance over any individual classification method in these scenarios. Thus, strategies that combine the predictions of different classifiers may prove to be optimal in currently ongoing and forthcoming photometric surveys, such as the Dark Energy Survey and the Large Synoptic Survey Telescope.

preprint2014arXiv

Dark aspects of massive spinor electrodynamics

We investigate the cosmology of massive spinor electrodynamics when torsion is non-vanishing. A non-minimal interaction is introduced between the torsion and the vector field and the coupling constant between them plays an important role in subsequential cosmology. It is shown that the mass of the vector field and torsion conspire to generate dark energy and pressureless dark matter, and for generic values of the coupling constant, the theory effectively provides an interacting model between them with an additional energy density of the form $\sim 1/a^6$. The evolution equations mimic $Λ$CDM behavior up to $1/a^3$ term and the additional term represents a deviation from $Λ$CDM. We show that the deviation is compatible with the observational data, if it is very small. We find that the non-minimal interaction is responsible for generating an effective cosmological constant which is directly proportional to the mass squared of the vector field and the mass of the photon within its current observational limit could be the source of the dark energy.

preprint2013arXiv

Chaotic dynamics of the Bianchi IX universe in Gauss-Bonnet gravity

We investigate the dynamics of closed FRW universe and anisotropic Bianchi type-IX universe characterized by two scale factors in a gravity theory including a higher curvature (Gauss-Bonnet) term. The presence of the cosmological constant creates a critical point of saddle type in the phase space of the system. An orbit starting from a neighborhood of the separatrix will evolve toward the critical point, and it eventually either expands to the de Sitter space or collapses to the big crunch. In the closed FRW model, the dynamics is reduced to hyperbolic motions in the two-dimensional center manifold, and the system is not chaotic. In the anisotropic model, anisotropy introduces the rotational mode, which interacts with the hyperbolic mode to present a cylindrical structure of unstable periodic orbits in the neighborhood of the critical point. Due to the non-integrability of the system, the interaction of rotational and hyperbolic modes makes the system chaotic, making it impossible for us to predict the final fate of the universe. We find that the chaotic dynamics arises from the fact that orbits with even small perturbations around the separatrix oscillate in the neighborhood of the critical point before finally expanding or collapsing. The chaotic character is also evidenced by the fractal structures in the basins of attraction.