Artificial intelligence in the detection of Radio Galaxies.
Abstract
The article deals with artificial intelligence and machine learning, applying two reference automatic learning algorithms: the decision tree and the backward propagation neural network implemented in 30% of the Galaxy data set to observe and analyze the classification performance of machine learning techniques. A sample of radio sources with a radio brightness of 1.4 GHz, S1.4> 10mJy, which employs a new technique more efficient than previous methods, such as a pronounced spectral index or a small angular size, using near-infrared data to filter at low, they use radiogalaxies with redshift (z <2) by including only sources with a very weak identification or no detection in the K band or 3.6 μm.