Using Machine Learning and Light Spatial Sequence Arrangement for Copying Positioning Unit Cell to Reduce Training Burden in Visible Light Positioning (VLP).