Single Training Dimension Selection for Word Embedding with PCA
This incremental method helps researchers and practitioners in NLP by reducing computational costs for training embeddings for tasks like sentiment analysis and question answering.
The paper tackles the problem of selecting the optimal number of dimensions for word embeddings by proposing a PCA-based method that trains a single high-dimensional embedding and incrementally removes dimensions while monitoring performance, achieving a 10x reduction in training sets while maintaining optimal accuracy.
In this paper, we present a fast and reliable method based on PCA to select the number of dimensions for word embeddings. First, we train one embedding with a generous upper bound (e.g. 1,000) of dimensions. Then we transform the embeddings using PCA and incrementally remove the lesser dimensions one at a time while recording the embeddings' performance on language tasks. Lastly, we select the number of dimensions while balancing model size and accuracy. Experiments using various datasets and language tasks demonstrate that we are able to train 10 times fewer sets of embeddings while retaining optimal performance. Researchers interested in training the best-performing embeddings for downstream tasks, such as sentiment analysis, question answering and hypernym extraction, as well as those interested in embedding compression should find the method helpful.