Proto Planetary Disk Model Inversion Using Artificial Neural Networks

Details Proto Planetary Disk Model Inversion Using Artificial Neural Networks Proto Planetary Disk Model Inversion Using Artificial Neural Networks

Dec 2006

adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2006aas...209.7609r&link_type=abstract

2007 AAS/AAPT Joint Meeting, American Astronomical Society Meeting 209, #76.09; Bulletin of the American Astronomical Society, V

Astronomy and Astrophysics

Astronomy

Scientific paper

We explore the use of artificial neural networks (ANNs) both to increase the calculation speed of protoplanetary disk models and to rapidly recover model parameters from real observational data. The problem of recovering physical parameters from observational data is known as inversion. It is the inverse problem of modeling, which is the production of simulated observations from physical parameters. We use the two layer protoplanetary disk model of Dullemond et al. (2001) to create an ANN capable of quickly performing inversions. The model is used to calculate a grid of physical parameters versus the associated simulated spectra. The ANN is trained to mimic the grid, taking the simulated spectra on the input producing physical parameters on the output. From our point of view, the ANN can be viewed as multi-dimensional interpolation machine. It has the ability to generalize and interpolate between the grid points. In principle, real observational data can be presented to the ANN and it will quickly produce physical parameters at the output. We compare the speed of inversion and the recovered parameters by the ANN with the results of a common brute force method of performing inversions via least squares minimization.

Affiliated with

Also associated with

No associations

Rating

Proto Planetary Disk Model Inversion Using Artificial Neural Networks does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Proto Planetary Disk Model Inversion Using Artificial Neural Networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Proto Planetary Disk Model Inversion Using Artificial Neural Networks will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-1157899