Back

Speaker "Piotr Teterwak" Details Back

 

Topic

Deep Learning, Machine Learning

Abstract

We start out with a description of what deep learning is and give examples of applications where it’s been successful. Then we walk through the setup of the GraphLab Create package. In the second half of the tutorial, we walk through a IPython notebook that constructs an image similarity search application using a deep learning model and a nearest neighbors toolkit. If time allows, we will discuss how to combine image features and text features to improve the quality of search results. Folks should walk away knowing how to start building an image analysis application on their own using deep learning.

Profile

Piotr is a machine learning engineer at Dato (formerly known as GraphLab) developing tools and algorithms that make it straightforward to build, train, and deploy deep learning models. He is passionate about making machine learning tractable and easy to use. Piotr studied Computer Science at Dartmouth College, where he worked on deep learning in the context of computer vision.