From the course: TensorFlow: Practical Skills in Constructing, Training, and Optimizing Models
Join today to access over 25,300 courses taught by industry experts.
Explore advanced optimization techniques - TensorFlow Tutorial
From the course: TensorFlow: Practical Skills in Constructing, Training, and Optimizing Models
Explore advanced optimization techniques
- [Instructor] Probably your neural network won't perform optimally right off the bat. Different activation functions, different architectures, different optimizers and loss functions, different strengths of regularization, along with other parameter tweaks will work best for different sorts of machine learning problems. The number of possibilities can be bewildering but don't panic. In this lesson, I'll take you through a few common techniques to try to improve the speed and performance of your neural network. I'll start with some critical parameters we can tune inside of the dense layer class. Open up 0402_Explore.ipynb. Here I have imported TensorFlow as usual but I have also, for the sake of convenience, dug into the Keras library to grab the layer's module and the sequential model type. First of all, TensorFlow lets you choose the activation function you want to set for the nodes in each layer. After data and weight matrices have been manipulated, the activation function…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.