[Adapt] Seminar

黄姗姗 798508656 at qq.com
Tue Nov 10 22:55:18 CST 2020


Hi Adapters,


     Neural networks can achieve impressive performance on many natural language processing

applications, but they typically need large labeled data for training and are not easily interpretable. On the other hand, symbolic rules such as regular expressions are interpretable, require no training, and often achieve decent accuracy; but rules cannot benefit from labeled data when available and hence underperform neural networks in rich-resource scenarios. How to combine the advantages of symbolic rules and neural networks is an open question and is drawing increasing attention recently.




    In this seminar, I want to present a type of recurrent neural networks called FA-RNNs that combine the advantages of neural networks and regular expression rules.


Related papers:

Cold-Start and Interpretability: Turning Regular Expressions into Trainable Recurrent Neural Networks




Hope you can gain a fresh perspective after the talk.


Time: Wed 4:00pm

Venue: SEIEE 3-414

Best regards,

Shanshan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.sjtu.edu.cn/pipermail/adapt/attachments/20201110/81f07714/attachment.html>


More information about the Adapt mailing list