Bulletin of the American Physical Society
APS March Meeting 2019
Volume 64, Number 2
Monday–Friday, March 4–8, 2019; Boston, Massachusetts
Session A21: Big Data in PhysicsCareers Focus Undergraduate
|
Hide Abstracts |
Sponsoring Units: DCOMP Chair: Hans Herrmann, Universidade Federal do Ceará Room: BCEC 157B |
Monday, March 4, 2019 8:00AM - 8:36AM |
A21.00001: Big data in physics, biology and social science Invited Speaker: Flaviano Morone In the digital age, we are creating 2.5 billion gigabytes of data every single day of our lives, and the 75% is unstructured, coming from texts, videos and opinions spread through online social networks. These data sets are so voluminous and complex that traditional data analytics methods are often inadequate to extract value from them. |
Monday, March 4, 2019 8:36AM - 9:12AM |
A21.00002: AI Astrophysics Invited Speaker: Kevin Schawinski I present how some recent advances in artificial intelligence can help us explore astrophysical problems. Astrophysics faces a particular challenge from the fact that experiments with controlled variables are impossible. New approaches such as generative models can help us better constrain and understand the underlying processes governing the formation and evolution of galaxies and other objects. While cutting edge research from computer science can be helpful in many case, new tools abstracting AI/ML to a more user-friendly level also have a significant potential for advancing research in astrophysics, and other areas of the physical sciences. |
Monday, March 4, 2019 9:12AM - 9:48AM |
A21.00003: Insights into the world connectivity from aviation and tourism data Invited Speaker: Nuno Araujo The World Airline Network (WAN) is an infrastructure that reduces the geographical gap between societies, both small and large, and bring forth economic gains. With the extensive use of a publicly maintained data set that contains information about airports and alternative connections between these airports, we empirically reveal that WAN is a redundant and resilient network for long distance air travel, but otherwise breaks down completely due to removal of short and apparently insignificant connections [1]. These short-range connections with moderate number of passengers are the connections that keep remote parts of the world accessible. It is surprising, insofar as there exists a highly resilient and strongly connected core consisting of a small fraction of airports (around 2.3%) together with an extremely fragile star-like periphery. The core-periphery structure is also observed for a number of important transport networks. With a dynamic model for the evolution of transport networks, we show that core-periphery structures are (very likely) a product of a tug of war between connectivity and economic profit [2]. Finally, we will discuss how the mobility pattern of tourists around the world correlates with the underlying Airline Network [3]. |
Monday, March 4, 2019 9:48AM - 10:00AM |
A21.00004: Dynamics of Traffic Congestions from Large-Scale Data in the Taiwan Highway System T. S. Choi, Yee Man Tai, Yulin Xu, Kwong Tai Siu, Ki Wing To, K.Y. Michael Wong The Taiwan Highway System has the longest electronic toll collection (ETC) freeway mileage in the world. There are 300 sensors installed on the two North-South spanning highways. The collected data consists of the time information of individual vehicles passing through ETC sensors along their highway journeys. Highway segments are demarcated by successive sensors, enabling us to extract a tremendous volume of dynamical segment-wise data. We trace the evolution of the system on the so-called fundamental diagram with vehicle flux versus vehicle density during traffic congestions in segments, and show that congestions are characterized by loopy trajectories in the diagram. By considering the area enclosed by a loop, we find that there are two types of congestion dynamics -- moderate flow and serious congestion. They behvae differently in terms of the area enclosed. Data extracted from the time delays of individual vehicles show that the area enclosed is a measure of the economic loss due to congestion. |
Monday, March 4, 2019 10:00AM - 10:12AM |
A21.00005: How the brain transitions from conscious to subliminal perception Francesca Arese Lucini We study the transition in the functional network that characterize the human brains' conscious state to an unconscious subliminal state of perception by using k-core percolation. We find that the most inner core (ie, the most connected kernel) of the functional network in the state of consciousness (the visual cortex and the left middle frontal gyrus) is represented by the areas that remain active when the brain transitions to the subliminal unconscious state. That is, the inner core of the conscious network coincides with the unconscious state. Based on data analysis and mathematical modeling we interpret these results as a transition driven by k-core percolation, where the conscious state is inactivated by the disappearance of the peripheral shells in the k-shell decomposition structure. Thus, the inner core and most robust component of the brain seems to be the unconscious subliminal state. If this result were to be found valid for other states of the brain, it could set interesting constraints to models of consciousness and brain structure, in that the location of the core of the functional brain network is in the unconscious part of the brain rather than in the conscious state as previously thought. |
Monday, March 4, 2019 10:12AM - 10:24AM |
A21.00006: Super Resolution Convolutional Neural Network for Feature Extraction in Spectroscopic Data Han Peng, Xiang Gao, Yu He, Yiwei Li, Yuchen Ji, Chuhang Liu, Sandy Adhitia Ekahana, Ding Pei, zhongkai liu, Zhixun Shen, Yulin Chen Two dimensional (2D) peak finding is a common practice in data analysis for physics experiments, which is typically achieved by computing the local derivatives. However, this method is inherently unstable when the local landscape is complicated, or the signal-to-noise ratio of the data is low. In this work, we propose a new method in which the peak tracking task is formalized as an inverse problem, thus can be solved with a convolutional neural network (CNN). In addition, we show that the underlying physics principle of the experiments can be used to generate the training data. By generalizing the trained neural network on real experimental data, we show that the CNN method can achieve comparable or better results than traditional derivative based methods. This approach can be further generalized in different physics experiments when the physical process is known. |
Monday, March 4, 2019 10:24AM - 10:36AM |
A21.00007: A Library for Real Time Interactive High Performance Computing of 2D and 3D Physics Problems through GPU Flavio Fenton, Abouzar Kaboudian Many interesting and complex physics problems such as those that use large scale reaction-diffusion equations to study animal skin patterns, brain activity and cardiac arrhtyhmias require the use of supercomputers. Likewise is the case for systems displaying turbulent fluid dynamics, surface growth, and systems using heat and wave equations. We have developed a library using WebGL, which allows to code and run physics problems, that generally require supercomputers to run, in the GPU of a local PC and even a cellphone, which can run up to 7billion differential equations per second. |
Monday, March 4, 2019 10:36AM - 10:48AM |
A21.00008: A Robust Approach to Compressive Sensing by Shortest-Solution Decimation Mutian Shen, pan zhang, Haijun Zhou Compressed sensing is an important problem in many fields of science and engineering, including computational physics. It reconstructs signals by finding sparse solutions to underdetermined linear equations. We propose a deterministic and non-parametric algorithm, shortest-solution guided decimation (SSD), to construct support of the sparse solution under the guidance of the dense least-squares solution of the recursively decimated linear equation. The most significant feature of SSD is its insensitivity to correlations in the sampling matrix. Using extensive numerical experiments, we show that SSD greatly outperforms L1 -norm based methods, orthogonal least squares, orthogonal matching pursuit, and approximate message passing when the sampling matrix contains strong correlations. This nice property of correlation tolerance makes SSD a versatile and robust tool for different types of real-world signal acquisition tasks. |
Monday, March 4, 2019 10:48AM - 11:00AM |
A21.00009: TBTK: General purpose data structures for quantum mechanics Kristofer Björnson The scientific community has developed a wide range of algorithms and software packages for solving quantum mechanical problems. However, interfacing these methods with each other can be a difficult task. The main reason is a lack of cross application conventions. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700