Estimating the continuum of quasars using the arti cial neural networks

A lot of absorption lines are in the bluewards of Lyα emission line of quasar which is well-known as Lyαforest. Most of absorption lines in this forest belong to the Lyα absorption of the neutral hydrogen in the inter-galactic medium (IGM). For high redshift quasars and in the continuum with low and medium resolution, there are no many regions without absorption, so that, the quasar continuum in the forest is not obvious. Determination of the continuum in the forest is essential to study material distribution in the IGM, which is conductible through these absorption lines. One way to find this continuum is to predict it using longer wavelengths of the Lyα emission line of quasar, redwards of quasar continuum. Principal component analysis (PCA) method was proposed by researcher to estimate the bluewards of 50 low redshift quasars with 9% mean absolute error (error range was3-30%). In this article, the whole continuum is predicted using only the redwards of the quasar continuum and ten random data of the forest by an artificial neural network (ANN). Five different training algorithms are used to train the ANN. The simulation results show that mean absolute error for the Lyα forest is decreased to 5.27% (with error range between 1.63-9.05%). These results verify the capability of the ANN to predict the quasar continuum in the Lyαforest as compared with the statistical methods.