6:30 Networking and food (food sponsored by http://www.capestart.com)
7:00 Katherine Bailey, Principal Data Scientist at Acquia
Classifying Content Using Few-Shot Learning
Text classification is usually treated as a supervised learning problem: given a set of labeled texts, train a classifier such that it can accurately predict the classes of unseen texts. But many businesses have large amounts of text data where few or no texts have been classified. This talk will introduce a way of classifying an entire set of unlabeled texts using a human-in-the-loop approach, where the content owner manually classifies just one or two items per category and the rest can be automatically classified. The technique uses transfer learning (in the form of pre-trained word embeddings), human-in-the-loop ML and “few-shot” learning.
7:30 Sara Robinson, Developer Advocate at Google
Josh Gordon, Developer Advocate at Google
Building a Custom Text Classification Model With Keras
Most businesses deal with tons of text data on a daily basis. In this talk, we'll show you how to build a machine learning model that tags text into custom categories specific to your dataset. We'll build our model with Keras, a high-level API for TensorFlow. We'll start by giving an overview of TensorFlow and Keras, and then we'll dive into a live demo of our text classification model. 8:00 Networking and wrap up