Combining bag-of-words and sentiment features of annual reports to predict abnormal stock returns.
Petr HájekPublished in: Neural Comput. Appl. (2018)
Keyphrases
- bag of words
- textual features
- image classification
- late fusion
- visual words
- text classification
- n gram
- action recognition
- image features
- image representation
- bag of words models
- document representation
- stock returns
- feature vectors
- text classification tasks
- spatial pyramid matching
- feature set
- feature space
- sentiment classification
- feature extraction
- image processing
- machine learning
- co occurrence
- object detection
- region of interest
- sentiment analysis
- cross sectional