A Relative Comparison of Training Algorithms in Artificial Neural Network

A novel approach in predicting virtual garment fitting sizes with psychographic characteristics and 3D body measurements using artificial neural network and visualizing fitted bodies using generative adversarial network

“…The division for training and data testing is as follows: In the ANN training-testing procedure, 70 randomly selected respondents' data would serve as the training data and 50 respondents' data would serve as the testing data. The data was normalized into the format [0, 1], and the learning rate was set to the comparative effective value of 0.001 [ 42 , 43 ]. Two features, the optimizer and loss function, were defined before model training.…”