Главная
Study mode:
on
1
A Prediction Problem
2
Types of Prediction
3
Why Call it "Structured" Prediction?
4
Many Varieties of Structured Prediction!
5
Sequence Labeling w
6
Why Model Interactions in Output? . Consistency is important!
7
A Tagger Considering Output Structure movie
8
Training Structured Models
9
Local Normalization and
10
The Structured Perceptron Algorithm . An extremely simple way of training (non-probabilistic) global models . Find the one-best, and it's score is better than the correct answer adjust parameters to …
11
Contrasting Perceptron and Global Normalization
12
Structured Training and Pre-training
13
Hinge Loss for Any Classifier! We can swap cross-entropy for hinge loss anytime
14
Cost-augmented Hinge
15
Costs over Sequences
16
Cost-Augmented Decoding for Hamming Loss
17
Solution 1: Sample Mistakes in Training (Ross et al. 2010)
Description:
Explore structured prediction basics in this comprehensive lecture from CMU's Neural Nets for NLP 2018 course. Delve into various prediction types, understand the importance of modeling output interactions, and learn about sequence labeling. Discover training methods for structured models, including local normalization and the structured perceptron algorithm. Compare perceptron and global normalization approaches, and examine the use of hinge loss in classifiers. Investigate cost-augmented hinge loss, costs over sequences, and cost-augmented decoding for Hamming loss. Conclude by exploring solutions for sampling mistakes during training, as presented by Ross et al. in 2010.

CMU Neural Nets for NLP - Structured Prediction Basics

Graham Neubig
Add to list
#Computer Science #Artificial Intelligence #Neural Networks #Natural Language Processing (NLP) #Machine Learning #Sequence Labeling
1 of 17

A Prediction Problem