More Features Are Not Always Better: Evaluating Generalizing Models in Incident Type Classification of Tweets.
Axel SchulzChristian GuckelsbergerBenedikt SchmidtPublished in: EMNLP (2015)
Keyphrases
- classification models
- feature vectors
- feature extraction
- classification accuracy
- classification process
- pattern recognition
- machine learning approaches
- benchmark datasets
- machine learning
- feature set
- predictive power
- feature space
- prior knowledge
- machine learning models
- classification schemes
- image classification
- model selection
- classification method
- bayesian classifier
- feature selection
- class labels
- learned models
- pattern classification
- probabilistic model
- machine learning methods
- decision trees
- support vector
- extracted features
- svm classifier
- feature weights
- machine learning algorithms
- features extraction
- feature analysis
- neural network
- supervised learning
- hidden markov models
- feature selection algorithms
- feature subset
- support vector machine
- text classification
- image features