Data-Efficient Classification of Radio Galaxies
This work provides an automated classification technique for radio galaxies, which will be crucial for astronomers analyzing data from upcoming surveys with next-generation radio telescopes expected to detect hundreds of thousands of new radio galaxies.
This paper addresses the classification of radio galaxies into morphological classes (FRI, FRII, Bent, Compact) using deep learning on a small dataset of approximately 2000 samples. The best performing model achieved over 92% classification accuracy, with Bent and FRII galaxies being the primary source of confusion.
The continuum emission from radio galaxies can be generally classified into different morphological classes such as FRI, FRII, Bent, or Compact. In this paper, we explore the task of radio galaxy classification based on morphology using deep learning methods with a focus on using a small scale dataset ($\sim 2000$ samples). We apply few-shot learning techniques based on Twin Networks and transfer learning techniques using a pre-trained DenseNet model with advanced techniques like cyclical learning rate and discriminative learning to train the model rapidly. We achieve a classification accuracy of over 92\% using our best performing model with the biggest source of confusion being between Bent and FRII type galaxies. Our results show that focusing on a small but curated dataset along with the use of best practices to train the neural network can lead to good results. Automated classification techniques will be crucial for upcoming surveys with next generation radio telescopes which are expected to detect hundreds of thousands of new radio galaxies in the near future.