Hierarchical sparse autoencoder using linear regression-based features in clustering for handwritten digit recognition.
Hai T. PhanAn T. DuongSon T. TranPublished in: ISPA (2013)
Keyphrases
- handwritten digit recognition
- hierarchical clustering
- machine learning
- clustering algorithm
- feature extraction
- low level
- clustering method
- similarity function
- unsupervised feature selection
- translation and rotation invariant
- neural network
- eigenvalue decomposition
- additional features
- sparse data
- high dimensionality
- feature set
- image features