Optimization of Fuzzy System Inference Model on Mini Batch Gradient Descent

Optimization is one of the factors in machine learning to help model training during backpropagation. This conducted by adjusting weights minimize loss function and overcome dimensional problems. Also, gradient descent method a simple approach backpropagation solve minimum The mini-batch (MBGD) methods proven be powerful for large-scale learning. addition several approaches MBGD such as AB, BN, UR can accelerate convergence process, hence, algorithm becomes faster more effective. added will perform an optimization process on results data rule that has been processed its objective function. processing showed MBGD-AB-BN-UR stable computational time three sets than other methods. For evaluation, this research used RMSE, MAE, MAPE.