SAS Tutorial | How to use Dropout in Deep Learning

In this SAS How To Tutorial, Robert Blanchard takes a look at using drop out in deep learning. Dropout may be used to make a deep learning model more generalizable. Robert uses real-world examples to explain the concept of dropout and how it helps in the regularization of a deep learning model. After a brief discussion of deep learning, Roberts steps through the process of building a denoising convolutional autoencoder where dropout is added to a deep learning model using SAS Studio.

Download Data Files

Chapters
0:00 – What is Dropout in Deep Learning
1:11 – Example of how to use Dropout in Deep Learning using SAS Studio
3:40 – Pro tip discussion: Randomly shuffle the data

Learn more about SAS Software

SUBSCRIBE TO THE SAS USERS YOUTUBE CHANNEL #SASUsers #LearnSAS

ABOUT SAS
SAS is a trusted analytics powerhouse for organizations seeking immediate value from their data. A deep bench of analytics solutions and broad industry knowledge keep our customers coming back and feeling confident. With SAS®, you can discover insights from your data and make sense of it all. Identify what’s working and fix what isn’t. Make more intelligent decisions. And drive relevant change.

CONNECT WITH SAS

[marcellosandipinheiro]

It was very good Robert. Lined domain knowledge brilliantly, so that it is easy to learn.
thank you

[itcu]

So what platform are you using it doesn't look like base sas, is it just the university edition?

[LaSupp]

Thanks for the content Robert. Your mentioned you model has around 170, 000 parameters. Can you help me understand what these parameters are? Is your data structured as one row per image? If so, would it be the case that every pixel is a parameter?