Details
| Original language | English |
|---|---|
| Article number | 117 |
| Number of pages | 12 |
| Journal | Astrophysical Journal |
| Volume | 781 |
| Issue number | 2 |
| Early online date | 16 Jan 2014 |
| Publication status | Published - 1 Feb 2014 |
Abstract
In the modern era of big data, many fields of astronomy are generating huge volumes of data, the analysis of which can sometimes be the limiting factor in research. Fortunately, computer scientists have developed powerful data-mining techniques that can be applied to various fields. In this paper, we present a novel artificial intelligence (AI) program that identifies pulsars from recent surveys by using image pattern recognition with deep neural nets - the PICS (Pulsar Image-based Classification System) AI. The AI mimics human experts and distinguishes pulsars from noise and interference by looking for patterns from candidate plots. Different from other pulsar selection programs that search for expected patterns, the PICS AI is taught the salient features of different pulsars from a set of human-labeled candidates through machine learning. The training candidates are collected from the Pulsar Arecibo L-band Feed Array (PALFA) survey. The information from each pulsar candidate is synthesized in four diagnostic plots, which consist of image data with up to thousands of pixels. The AI takes these data from each candidate as its input and uses thousands of such candidates to train its ;9000 neurons. The deep neural networks in this AI system grant it superior ability to recognize various types of pulsars as well as their harmonic signals. The trained AI's performance has been validated with a large set of candidates from a different pulsar survey, the Green Bank North Celestial Cap survey. In this completely independent test, the PICS ranked 264 out of 277 pulsar-related candidates, including all 56 previously known pulsars and 208 of their harmonics, in the top 961 (1%) of 90,008 test candidates, missing only 13 harmonics. The first non-pulsar candidate appears at rank 187, following 45 pulsars and 141 harmonics. In other words, 100% of the pulsars were ranked in the top 1% of all candidates, while 80% were ranked higher than any noise or interference. The performance of this system can be improved over time as more training data are accumulated. This AI system has been integrated into the PALFA survey pipeline and has discovered six new pulsars to date.
Keywords
- pulsars: general, stars: neutron, techniques: image processing, words: methods: data analysis
ASJC Scopus subject areas
- Physics and Astronomy(all)
- Astronomy and Astrophysics
- Earth and Planetary Sciences(all)
- Space and Planetary Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Astrophysical Journal, Vol. 781, No. 2, 117, 01.02.2014.
Research output: Contribution to journal › Review article › Research › peer review
}
TY - JOUR
T1 - Searching for pulsars using image pattern recognition
AU - Zhu, W. W.
AU - Berndsen, A.
AU - Madsen, E. C.
AU - Tan, M.
AU - Stairs, I. H.
AU - Brazier, A.
AU - Lazarus, P.
AU - Lynch, R.
AU - Scholz, P.
AU - Stovall, K.
AU - Ransom, S. M.
AU - Banaszak, S.
AU - Biwer, C. M.
AU - Cohen, S.
AU - Dartez, L. P.
AU - Flanigan, J.
AU - Lunsford, G.
AU - Martinez, J. G.
AU - Mata, A.
AU - Rohr, M.
AU - Walker, A.
AU - Allen, B.
AU - Bhat, N. D.R.
AU - Bogdanov, S.
AU - Camilo, F.
AU - Chatterjee, S.
AU - Cordes, J. M.
AU - Crawford, F.
AU - Deneva, J. S.
AU - Desvignes, G.
AU - Ferdman, R.
AU - Freire, P. C.C.
AU - Hessels, J. W.T.
AU - Jenet, F. A.
AU - Kaplan, D. L.
AU - Kaspi, V. M.
AU - Knispel, B.
AU - Lee, K. J.
AU - Van Leeuwen, J.
AU - Lyne, A. G.
AU - McLaughlin, M. A.
AU - Siemens, X.
AU - Spitler, L. G.
AU - Venkataraman, A.
PY - 2014/2/1
Y1 - 2014/2/1
N2 - In the modern era of big data, many fields of astronomy are generating huge volumes of data, the analysis of which can sometimes be the limiting factor in research. Fortunately, computer scientists have developed powerful data-mining techniques that can be applied to various fields. In this paper, we present a novel artificial intelligence (AI) program that identifies pulsars from recent surveys by using image pattern recognition with deep neural nets - the PICS (Pulsar Image-based Classification System) AI. The AI mimics human experts and distinguishes pulsars from noise and interference by looking for patterns from candidate plots. Different from other pulsar selection programs that search for expected patterns, the PICS AI is taught the salient features of different pulsars from a set of human-labeled candidates through machine learning. The training candidates are collected from the Pulsar Arecibo L-band Feed Array (PALFA) survey. The information from each pulsar candidate is synthesized in four diagnostic plots, which consist of image data with up to thousands of pixels. The AI takes these data from each candidate as its input and uses thousands of such candidates to train its ;9000 neurons. The deep neural networks in this AI system grant it superior ability to recognize various types of pulsars as well as their harmonic signals. The trained AI's performance has been validated with a large set of candidates from a different pulsar survey, the Green Bank North Celestial Cap survey. In this completely independent test, the PICS ranked 264 out of 277 pulsar-related candidates, including all 56 previously known pulsars and 208 of their harmonics, in the top 961 (1%) of 90,008 test candidates, missing only 13 harmonics. The first non-pulsar candidate appears at rank 187, following 45 pulsars and 141 harmonics. In other words, 100% of the pulsars were ranked in the top 1% of all candidates, while 80% were ranked higher than any noise or interference. The performance of this system can be improved over time as more training data are accumulated. This AI system has been integrated into the PALFA survey pipeline and has discovered six new pulsars to date.
AB - In the modern era of big data, many fields of astronomy are generating huge volumes of data, the analysis of which can sometimes be the limiting factor in research. Fortunately, computer scientists have developed powerful data-mining techniques that can be applied to various fields. In this paper, we present a novel artificial intelligence (AI) program that identifies pulsars from recent surveys by using image pattern recognition with deep neural nets - the PICS (Pulsar Image-based Classification System) AI. The AI mimics human experts and distinguishes pulsars from noise and interference by looking for patterns from candidate plots. Different from other pulsar selection programs that search for expected patterns, the PICS AI is taught the salient features of different pulsars from a set of human-labeled candidates through machine learning. The training candidates are collected from the Pulsar Arecibo L-band Feed Array (PALFA) survey. The information from each pulsar candidate is synthesized in four diagnostic plots, which consist of image data with up to thousands of pixels. The AI takes these data from each candidate as its input and uses thousands of such candidates to train its ;9000 neurons. The deep neural networks in this AI system grant it superior ability to recognize various types of pulsars as well as their harmonic signals. The trained AI's performance has been validated with a large set of candidates from a different pulsar survey, the Green Bank North Celestial Cap survey. In this completely independent test, the PICS ranked 264 out of 277 pulsar-related candidates, including all 56 previously known pulsars and 208 of their harmonics, in the top 961 (1%) of 90,008 test candidates, missing only 13 harmonics. The first non-pulsar candidate appears at rank 187, following 45 pulsars and 141 harmonics. In other words, 100% of the pulsars were ranked in the top 1% of all candidates, while 80% were ranked higher than any noise or interference. The performance of this system can be improved over time as more training data are accumulated. This AI system has been integrated into the PALFA survey pipeline and has discovered six new pulsars to date.
KW - pulsars: general
KW - stars: neutron
KW - techniques: image processing
KW - words: methods: data analysis
UR - http://www.scopus.com/inward/record.url?scp=84892723868&partnerID=8YFLogxK
U2 - 10.48550/arXiv.1309.0776
DO - 10.48550/arXiv.1309.0776
M3 - Review article
AN - SCOPUS:84892723868
VL - 781
JO - Astrophysical Journal
JF - Astrophysical Journal
SN - 0004-637X
IS - 2
M1 - 117
ER -