No
Yes
View More
View Less
Working...
Close
OK
Cancel
Confirm
System Message
Delete
Schedule
An unknown error has occurred and your request could not be completed. Please contact support.
Scheduled
Scheduled
Wait Listed
Personal Calendar
Speaking
Conference Event
Meeting
Interest
There aren't any available sessions at this time.
Conflict Found
This session is already scheduled at another time. Would you like to...
Loading...
Please enter a maximum of {0} characters.
{0} remaining of {1} character maximum.
Please enter a maximum of {0} words.
{0} remaining of {1} word maximum.
must be 50 characters or less.
must be 40 characters or less.
Session Summary
We were unable to load the map image.
This has not yet been assigned to a map.
Search Catalog
Reply
Replies ()
Search
New Post
Microblog
Microblog Thread
Post Reply
Post
Your session timed out.
This web page is not optimized for viewing on a mobile device. Visit this site in a desktop browser to access the full set of features.
Uppercase Letter
Lowercase Letter
Uppercase or Lowercase Letter
Number
Special Character
Password length of
or more and have
of the following:
Password 2 does not match password.
2018 GTC Washington DC
Add to My Interests
Remove from My Interests

DC8217 - Human + Machine: CNN and LSTM Based Approaches For Document Classification (Presented by Lockheed Martin)

Session Speakers
Session Description

We will discuss our experience developing and implementing a Human-in-the-Loop system for automatically classifying text documents with the goal of using very little training data and limiting human interaction. We will cover our journey through this space starting with a brief review of the Word2Vec algorithm, a foundational element of transfer learning for NLP. We will describe our implementation of a CNN architecture which resulted in near human performance, addressing the architecture decisions and tradeoffs. Finally, we will cover recent advancements in AWD LSTM-based transfer learning capabilities pioneered by Jeremy Howard and Sebastian Ruder at fast.ai and explain how these techniques address some of the tradeoffs from our previous model.


Additional Information
Deep Learning and AI
Deep Learning and AI
Aerospace, Defense, Manufacturing
Intermediate technical
Sponsor Session
50 minutes
Session Schedule
    Do Not Sell My Personal Information
    First name
    Last name
    Email address