Efficient similarity search based on data distribution properties in high dimensions

Full text