Structured Linear Models
author:
Fernando Pereira,
Instituto Superior Tècnico
Description
Over the last five years, we have been able to extend the theory of linear classifiers to structure prediction problems, combining the benefits of discriminative learning and of structured probabilistic models like hidden Markov models. I will review these models and their learning algorithms, and exemplify their use in text processing, with a focus on information extraction from biomedical text.
You might be experiencing some problems with Your Video player.
| Slides | |
| 0:00 | Structured Linear Models |
| 1:10 | Goals |
| 3:18 | Information Extraction |
| 5:06 | Biomedical Examples |
| 7:04 | Approach |
| 8:20 | Annotation Tool |
| 10:00 | Analyzing Text |
| 10:32 | Structured Classification |
| 11:45 | Challenges |
| 11:51 | Structured Classification (a) |
| 11:59 | Challenges (a) |
| 13:33 | Analysis by Tagging |
| 14:50 | Segmentation as Tagging |
| 15:43 | Traditional Approaches |
| 17:10 | Hidden Markov Model |
| 19:05 | HMMs in IE |
| 19:58 | Problems with HMMs |
| 20:03 | Hidden Markov Model (a) |
| 20:15 | Problems with HMMs (a) |
| 20:56 | Generating Multiple Features |
| 21:31 | Structured Linear Models |
| 25:06 | Learning |
| 25:17 | Structured Linear Models (a) |
| 25:32 | Learning (a) |
| 28:55 | Margin |
| 30:04 | Losses |
| 34:17 | Why? |
| 35:15 | Probabilistic Version |
| 36:15 | Features |
| 36:36 | Probabilistic Version (a) |
| 36:43 | Features (a) |
| 37:44 | MALLET |
| 39:05 | Evaluation |
| 39:44 | Gene/Protein Results |
| 41:50 | Variation Results |
| 42:32 | Tagger |
| 43:18 | Fable |
| 45:28 | Technical Challenges |
| 47:32 | Alternative: Online Training |
| 48:55 | Online Maximum Margin |
| 51:45 | Lists and Unlabeled Text pt 2 |
| 52:46 | Pattern Induction |
| 53:18 | Person Names |
| 53:53 | Improving CRF Tagger |
| 55:13 | Extensions |
| 63:05 | Losses (a) |
| 65:13 | Learning (b) |
Lecture rating
| People found this lecture: | ||
| Worth seeing | ||
| because it is: | ||
| Valuable and informative | ||
| Well presented | ||
| Easily understandable | ||
| Acceptably recorded | ||
| You need to login to cast your vote. | ||
Report a problem or upload files
If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Related content
Visitors who watched this lecture also watched...
34 comments
Download slides:
Link this page
Would you like to put a link to this lecture on your homepage?Go ahead! Copy the HTML snippet !
