A Comparative Study on Different Training Model in Machine Learning

Image reconstruction is the process of regeneration of an image from its distorted version. The image reconstruction problem is part of a larger field of image processing that is used in a variety of applications, including Forensic Analysis, Medical Science, Astronomy, Entertainment industry, Legal Services, etc. Designing a neural network that memorizes a given pattern or image and reconstructing the image requires resources, which is a tedious task and time-consuming. Many studies have been conducted in order to reconstruct images that have been distorted. This paper conducts a review of the current research to reconstruct the image from the distorted image using different Machine Learning methods. The domain of machine learning has led to the tremendous development of deep neural network techniques. Neural Networks are used to model complex patterns and predictive problems. This survey has compared different approaches for solving the problem and has proposed the neuromodulation technique as an efficient technique.