Improving Deep Neural Networks - Hyperparameter Tuning, Regularization and Optimization Coursera

Overview

By the end of the Improving Deep Neural Networks - Hyperparameter Tuning, Regularization and Optimization course offered by Coursera in partnership with Deeplearning, you will learn the best practices to train and develop test sets and analyze bias/variance for building deep learning applications; be able to use standard neural network techniques such as initialization, L2 and dropout regularization, hyperparameter tuning, batch normalization, and gradient checking; implement and apply a variety of optimization algorithms, such as mini-batch gradient descent, Momentum, RMSprop and Adam, and check for their convergence; and implement a neural network in TensorFlow.

Key facts

• The Deep Learning Specialization is our foundational programme that will help you understand the capabilities, challenges, and consequences of deep learning and prepare you to participate in the development of leading-edge AI technology. It provides a pathway for you to gain the knowledge and skills to apply machine learning to your work, level up your technical career, and take the definitive step in the world of AI.