Sumários
Information Extraction (cont.)
22 outubro 2018, 17:00 • Bruno Emanuel Da Graça Martins
- Supervised Learning of HMMs
- Improving Probability Estimates and Smoothing
- Unsupervised Learning of HMMs and the Baum-Welsh Method
- Other Sequential Classification Models
- Restructuring HMMs With Features
- Structured Perceptrons
- Linear-Chain Conditional Random Fields
Hidden Markov Models for Information Extraction (cont.)
19 outubro 2018, 14:00 • Pável Pereira Calado
- Supervised learning of HMMs
- Unsupervised learning of HMMs (the Segmental K-Means algorithm)
- A brief introduction to other sequential classification models for IE
- The structured Perceptron
- Conditional Random Fields
- Introduction to the course project
Text Classification and Clustering
19 outubro 2018, 11:00 • Danielle Caled Vieira
- Implementation of classifiers using Multinomial Naive Bayes, K-Neighbors, Perceptron and Support Vector Machines (scikit-learn)
- Stopwords removal
- Rare words removal
- Frequent words removal
- Implementation of K-means for clustering
- Performance evaluation
- Pen and paper exercise:
- Binary Naive Bayes
- Perceptron Classifier
- K-means algorithm
Text Classification and Clustering
19 outubro 2018, 09:30 • Danielle Caled Vieira
- Implementation of classifiers using Multinomial Naive Bayes, K-Neighbors, Perceptron and Support Vector Machines (scikit-learn)
- Stopwords removal
- Rare words removal
- Frequent words removal
- Implementation of K-means for clustering
- Performance evaluation
- Pen and paper exercise:
- Binary Naive Bayes
- Perceptron Classifier
- K-means algorithm
Information Extraction with Hidden Markov Models (cont.)
19 outubro 2018, 08:00 • Bruno Emanuel Da Graça Martins
- The backward procedure for estimating the probability of a sequence of states
- Computing the probability of a sequence of states and the Viterbi decoding algorithm
- The posterior decoding algorithm