Monday, May 1, 2017

Getting started with machine learning with Caffe

Many data centres are being touched by machine learning, as organisations increasingly seek out opportunities to use it for understanding their customers better, designing improved products, or automating manual processes. Here’s a tutorial that will quickly bring you up to speed on Caffe, a deep learning framework from the Berkeley Vision and Learning Center. It’s integrated with the Intel® Math Kernel Library 2017 to deliver optimal performance.

The tutorial takes you from installation, through a discussion of the data types and dataset preparation, into training the model, fine-tuning and testing. There’s also a guide to using the Python API for Caffe, which is handy in testing, classifying and feature extraction.  If you can’t get it working, check out the debugging guide.

For more resources to help you get the best out of your hardware, take a look at the machine learning and modern code sections of Intel® Developer Zone.