shakib-io / cse445_machine_learning Goto Github PK
View Code? Open in Web Editor NEWThis repo contains all of the files, slides, notes for the Machine Learning (CSE445) course. Denoising using an autoencoder is mainly a modification on the network to prevent it from learning the identity function. The autoencoder sometimes becomes so big that it only learns the data and makes the output as input. It doesn’t perform any dimension reduction rather corrupts the input data by adding some noise or mask in the input values.